Frequently Asked Questions

Yes. Operators recall presets in one click, run VMS-driven patrol tours, and build or edit tours themselves in the tour editor. Radha can also send a camera to a preset or start a tour by name.

Yes. Active Directory and LDAP sign-in are configured from Settings, alongside SAML and OIDC single sign-on, and every login is audit-logged.

Yes. Visylix takes a closer look at every part of a high-resolution frame, so small, distant people and objects that a single pass would miss are still found.

Yes. Floor-plan maps show live camera status and alarm badges, a geographic map view runs on your own map tiles, an operator console handles acknowledging, escalating and closing incidents, and a device-health console flags failing cameras and storage runway before anyone needs the footage.

Yes. A dry run shows exactly what a rule would have done against real history, with no side effects, before you arm it. Every real firing is traced too, so you can see afterwards exactly which actions ran and why.

Yes. Ask an open question such as "what happened at the loading dock this afternoon?" and Radha runs a multi-step investigation on its own. It pulls together people sightings, AI detections, system events, alerts and video search into findings with a time-ordered evidence timeline, and can turn the result into a printable incident report.

Yes. Describe what you are looking for, such as "a man in a red jacket near the entrance" or "a white van this morning", and Radha returns ranked, time-stamped matches you can play instantly. The search runs entirely on your own servers; no footage or query ever leaves your network.

Yes. Radha can look at any live camera or snapshot and describe the scene ("what is happening on the office camera right now?"), and it accepts spoken questions through fully local voice input. Both run on your own hardware, with no cloud vision or speech service involved.

Yes, and much of it happens automatically. When an operator marks an alert as a false alarm, that camera's threshold for PPE, object, fire and weapon detection tightens automatically, never below the operator's own setting, with one switch to turn it off. Dismissing an alert with a one-tap reason (false alarm, fog or steam, known person, duplicate) also feeds the effectiveness board, which proposes retunes for noisy sources that an operator approves in one click. Motion, audio, crowd and tamper detection each learn their own camera's normal, and anomaly detection learns each scene. Everything stays on your own servers.

Yes. A rule can combine a video condition with an access-control event. For example, a person seen at a door together with a denied badge at that same door becomes one forced-entry or tailgating alert. Sudarshan uses the door-to-camera mapping you already set up for access control, and everything runs on your own servers.

Yes. Pick a person from any detection and Visylix finds them by appearance on your other cameras within a time window, then lays out a chronological journey with travel times between cameras. Operators review and confirm each sighting in one click. Vehicle journeys are exact, because a number plate is a precise identifier.

Yes. Rules combine conditions with ANY (OR) or ALL (AND), and an ALL rule fires only when every condition happens on the same camera within a time window. Actions include lighting up the tile, notifying a channel, starting a recording, calling a webhook, SIEM, PSIM or MQTT, raising an alarm incident, talkdown through the camera speaker, and switching siren and door-strike relays. Physical actions are armed deliberately, and you can require a second operator to confirm before a rule can drive a relay, so the alert always fires but the physical action follows a human decision.

Yes. Radha learns each site's vocabulary: tell it once to "call this camera the gate cam" and every future conversation understands the nickname. It remembers each user's preferences, and feedback and corrections are kept on your own servers as a record your team controls.

No. Radha runs entirely on premise: its language model, scene description and voice input all run on your own hardware. Radha has no internet access by design and works fully air-gapped, so it suits government sites, defence installations and classified networks without modification.

No. Every detection, watch and overlay is processed on your own servers. No frames, detections or watches go to any outside service, so Sudarshan works fully air-gapped. Overlays arrive in under a second and reach every operator watching the same camera at once.

Yes, per camera and on your own servers. Anomaly detection learns each scene, and tamper, motion, crowd and audio detection each learn their own camera's normal. When an operator marks an alert as a false alarm, that camera's threshold for PPE, object, fire and weapon detection tightens automatically, never below the operator's own setting, with one switch to turn it off. The effectiveness board also proposes retunes an operator approves. Nothing leaves the premises, and every adjustment can be reversed.

Yes. Line Crossing Detection produces in, out, and net people counts per line with hourly breakdowns on a dashboard, not just individual crossing events, the directional counting product that retail and facilities teams expect.

Yes. Queue and Wait-Time Analytics reports the current line length, the live service rate (how fast the queue is actually clearing), and a predicted wait time for someone joining the queue now, computed from that clearing rate, alongside average and maximum wait. It alerts when a line grows past your threshold so staff can open another counter before customers leave.

No. Visylix AI runs fully on premise. Detection, learning, face watchlists and cross-camera search all run on your own hardware and work fully air-gapped. Your video never leaves your building.

Radha checks the user's permissions, camera access and licence on every tool call, and writes every action to the audit trail. It has no delete tool, so nothing can be removed through it. Its 13 high-impact actions, such as arming rules, restarting services and bulk camera changes, always ask for a confirmation that names the exact target, and the model cannot approve its own actions. Person searches based on race, ethnicity or religion are refused, and attempts to trick Radha with hidden instructions are blocked and logged.

AI runs only on the cameras you choose, with detection zones per camera and per analytic, and overlays appear only on cameras configured for AI. Everything is processed on your own servers, and AI analytics are billed per camera per month on the cameras you choose.

Sudarshan scores every alert for dispatch priority and measures which alert sources actually earn an operator response, then suggests retunes for the noisy ones. A flood of alerts becomes a ranked, actionable queue.

A rule that looks configured but never fires is the classic failure of automation. Sudarshan prevents it by design: the rule builder only offers a condition when the analytic behind it is enabled on every selected camera, and any rule that has never fired, has an unarmed physical action, or depends on an analytic that is no longer enabled appears on the rule health board. A dead rule is caught, never hidden.

From video alone, Visylix counts how many people pass through a secured door during an access event and flags group crossings. Connect a supported access-control system (Brivo, LenelS2, HID, AMAG or Genea/Openpath, or any other through the generic connector) and Visylix compares the number of people who crossed with the number of badges presented, alerting only when more people entered than credentials. That is the true tailgating signal.

Most systems either draw boxes on every camera until the wall is unreadable, or hide AI behind a separate analytics tab. Sudarshan puts live overlays only where the operator is looking and a quiet severity ring everywhere else. The live picture always comes first: overlays are a light, translucent layer that never covers or slows the video, and old overlay data fades out rather than misleading anyone.

Radha has 111 tools covering every part of Visylix: live view, recordings, snapshots, playback and PTZ; answers from every AI analytic; video search by description; multi-step investigations; watches and Sudarshan rules with dry runs; camera fault diagnosis and bulk actions; server capacity questions; recorder playback; PTZ presets and patrol tours; the alarm console, sensors, notifications and talkdown; daily briefings and incident reports; evidence locks; and administration including camera onboarding, users, face and plate enrolment, zones, sites, floor plans, integrations and licensing. It answers in English, Hindi and Hinglish, and it has no delete tool.

Sudarshan watches for specific faces, specific plates, plate patterns, a type of object inside a named zone (with a minimum count and dwell time), line crossings in a chosen direction, and anything you can describe in words. You describe what you want and Sudarshan arms a live watch that flags the moment it happens on the wall.

Sudarshan Rules is the automation builder inside Sudarshan. Where a watch flags a moment for an operator, a rule acts on it automatically, around the clock. You build a rule as IF something happens, WHEN a schedule applies, THEN take an action: for example, IF a person is detected on the loading dock, WHEN it is after 22:00, THEN light up the tile, notify the guard, sound the siren and start recording. It is built for the people who run the video wall, with no scripting.

Sudarshan is Visylix Live Intelligence: real-time AI on your video wall that matches detail to the operator's attention. The feed an operator is focused on gets rich live overlays (people, vehicles, faces, plates, zones and lines), while every other tile stays clean and lights up only when something needs attention. Operators can watch for anything by describing it, and alerts arrive in context. It is named after the Sudarshan Chakra, the ever-vigilant, all-seeing discus.

Cloud VMS vendors run their AI assistants in their own cloud; most on-premise vendors offer a search box at most. Radha is a complete copilot, covering conversation, search, investigation, scene description, voice, reports and administration, and it runs entirely on your own servers. It answers in English, Hindi and Hinglish, remembers the vocabulary your team teaches it, and is governed by confirmation on high-impact actions and a full audit trail.

Sudarshan draws live overlays for twelve analytics on the focused feed: object detection, person tracking, face recognition, number plates, pose, PPE, intrusion zones, line crossing, crowd counting, fire and smoke, weapons (guns and knives) and abandoned objects. Weapon boxes appear only once a weapon is confirmed across frames. Analytics without a box, such as audio events (gunshot, glass break, scream), anomaly detection, camera tampering and tailgating, signal on the same wall with a tile ring, an alert rail entry and one-click drill-in.

AI analytics are a separate subscription, billed per camera per month, on the cameras you choose. Pick any of the 22 analytics for each camera. Starter and Pro carry no AI cost in the plan. Talk to us for AI pricing.

Yes. Radha answers in English, Hindi and Hinglish, so an operator can ask "kal raat Gate 2 pe koi laal gaadi aayi thi kya?" and get the clips and the plate journey back. It also accepts spoken questions through fully local voice input.

Yes. PTZ autotracking uses live detections to keep a person or vehicle in frame as it moves, alongside one-click preset recall, patrol tours and a tour editor. Radha can recall a preset or start a tour by name.

It escalates. If nobody acknowledges a critical alert in time, Visylix escalates it to the next responder automatically, so a serious event is never left sitting on a screen.

Yes. Every redaction runs as a background job. Start a job, follow its progress and estimated time remaining, and cancel at any time. For long investigations you can queue many jobs and Visylix works through them in parallel.

Yes. Any share link can be revoked at any time, which blocks further access immediately and records the revocation in the audit log. Each share can also be limited to a number of views, and every access is logged.

Yes. Visylix works with no internet connection at all after installation. All AI analysis, video processing and management run locally.

Yes, and the deployment model is the reason. Visylix runs on infrastructure you control, including fully air-gapped, so personal data never has to leave your premises. That removes the cross-border transfer question rather than mitigating it. Visylix adds the tools privacy rules ask for: retention per camera, redaction on export, data export and erasure, consent records for face enrolment, and a tamper-evident audit trail. Visylix is engineered for the frameworks our buyers are audited against.

Yes. Visylix is hardware-agnostic and works with ONVIF and RTSP cameras from virtually any maker, so you can stage a migration instead of replacing an entire estate at once. Approved cameras are added alongside existing units, and NVR archive access keeps historical footage on legacy recorders reachable during the transition.

Yes. Once a user enrols an authenticator app or a passkey and enforcement is on, every sign-in requires the second factor, and repeated failures lock the account temporarily. For an organisation-wide policy, pair Visylix with single sign-on so your identity provider enforces two-factor sign-in for everyone.

Yes. Users provisioned through single sign-on have no local password, so they sign in only through your identity provider (SAML, OIDC, or Active Directory and LDAP), with group-to-role mapping and automatic account creation. A break-glass local administrator account is kept for emergencies, and every break-glass sign-in is audit-logged and can alert your security team.

Yes. Passkey sign-in happens entirely between the browser and your Visylix server, with no outside identity service, so it works fully air-gapped in any modern browser. Phone and laptop passkeys and hardware security keys are both supported.

Yes. Every action in Visylix is logged with time, user, source address and details, covering sign-ins, configuration changes, evidence access and every administrative operation. The audit trail is tamper-evident, so any alteration is detectable.

Yes. Visylix supports two-factor sign-in with an authenticator app and recovery codes on every paid plan, and passkeys from Pro upwards. Administrators can require it per user.

Yes. Visylix has four built-in roles, Super Admin, Admin, Operator and Viewer, mapped onto 46 permissions covering live view, recordings, camera settings, AI settings and system settings.

No. All 22 AI analytics and the Radha copilot run on your own hardware. Many platforms call themselves on premise while still sending video to a hosted service for analysis or for the assistant; Visylix does not, which is why it runs fully air-gapped.

NDAA Section 889 is mainly about equipment: cameras, recorders, encoders and network gear. Visylix is software that runs on your own servers and works with whichever approved cameras you select, so it adds no restricted hardware to your estate and lets you build a deployment that meets your procurement rules.

An operator creates a share link for a locked clip. The link carries a short-lived signed token, and the recipient opens it without a Visylix account and can view (or download, depending on policy) the clip until it expires. Every access is logged. Starter supports one recipient per share; Pro supports multiple named recipients, each with their own expiry and access log.

Face recognition is a policy decision rather than a default, and Visylix gives you the controls for it: consent recorded at enrolment, liveness-checked enrolment photos, erasure on request, retention rules and a full audit trail. Several jurisdictions treat biometric data as a special category, so check local rules. Where biometrics are not needed, intrusion, loitering, line crossing and crowd analytics answer most operational questions without processing faces at all.

Faces, plates and custom zones are redacted on export, so footage shared with police, insurers or the public shows only what it should. Access to each camera is controlled by permission, and every view and export is audit-logged.

Select a clip and choose a redaction preset: faces, number plates or both. Visylix finds every face or plate in the clip and blurs it, blurs any custom zones you draw exactly to their shape, and exports a standard MP4. Each redaction runs as a background job you can follow and cancel.

Repeated failed sign-in attempts lock the account for a period, and the lockout is tied to both the user and the source address, so an attacker cannot keep guessing. Every failure is audit-logged.

When a user signs in through single sign-on, Visylix reads the group claim from your identity provider and maps it to a Visylix role. A user in the "Security-Operators" group automatically becomes an operator. Roles update on every sign-in, so removing someone from a group removes their access at their next sign-in.

Minutes. Single sign-on is a short form in Settings: paste your identity provider's metadata address, map groups to Visylix roles and switch on automatic account creation. Visylix checks the connection before you turn it on, and there is no manual user sync.

Faces and plates are blurred as rectangles that follow each face or plate for as long as it is on screen. For whiteboards, screens or documents, you can draw custom zones, which are blurred exactly to the shape you draw. Redaction applies to exports.

Visylix gives you the tools GDPR asks for: fully on-premise deployment so data stays in your jurisdiction, retention rules per camera with automatic cleanup, redaction of faces, plates and custom zones on export, data export and erasure, consent records for face enrolment, and a tamper-evident audit trail. Visylix is engineered for the frameworks our buyers are audited against.

Visylix is hardware-agnostic software, so NDAA Section 889 questions attach to the cameras and recorders you deploy, not to the VMS. Visylix bundles no camera hardware and works with whichever approved cameras you select, so you can build an estate that meets your procurement rules.

Visylix runs on your own infrastructure, fully air-gapped if you need it, with encrypted connections, 46 permissions across 4 roles, single sign-on, Active Directory and LDAP, passkeys, retention rules, evidence locks sealed with SHA-256, and a tamper-evident audit trail of every operator action. Because it runs on your infrastructure, it sits inside your own audited environment. Visylix is engineered for the frameworks our buyers are audited against.

Visylix seals each locked recording with SHA-256 and keeps a full chain of custody. It exports a self-contained player that plays the clip offline, rechecks the seal in the viewer's browser and shows a readable timeline of custody events, so anyone can confirm the footage is untouched. It gives prosecutors and defence counsel exactly the evidence trail they ask for; admissibility is decided by the court.

No. Redaction is baked into the exported video, and there is no way to reverse it. That is deliberate: redacted footage is only trustworthy if it cannot be undone. The exported file can also be sealed, so any change after redaction is detectable.

Authenticator apps (Google Authenticator, Authy, 1Password and others) on Starter and above; passkeys and hardware security keys (YubiKey and others, plus iCloud Keychain, Windows Hello, Android and password managers) on Pro and above; and single-use recovery codes issued when two-factor sign-in is first enabled.

Every view, export and configuration change is written to a tamper-evident audit log against the operator who performed it. Recordings can be sealed with SHA-256 and locked so retention rules cannot delete them while an investigation is open. Logs can be streamed to a SIEM such as Splunk, Elastic or Datadog.

It stops a clip from being deleted by your retention rules, however old it gets. On Pro and above, locking also seals the clip with SHA-256, so any later change is detectable, and every access is recorded in a chain-of-custody trail. Locked clips export as a self-contained player that proves its own integrity offline.

Visylix encrypts every external connection, encrypts live video in the browser, and encrypts stored camera and system credentials. HTTPS is enforced across the dashboard, API and every management screen, and encryption keys stay on your own infrastructure.

Users receive single-use recovery codes when two-factor sign-in is first enabled, stored so that nobody, including us, can read them back. Administrators can also reset a user's two-factor sign-in, which triggers re-enrolment at the next sign-in. Every recovery-code use and reset is audit-logged.

Section 889 of the US National Defense Authorization Act restricts federal agencies, and anyone contracting with them, from buying or using certain telecommunications and video surveillance equipment. Part A bars the federal government from buying it. Part B bars the government from contracting with any organisation that uses that equipment anywhere in its operations.

Evidence management in Visylix covers bookmarks with headlines, notes, colours and tags; evidence locks sealed with SHA-256; a full chain of custody of who locked evidence, when and why; redaction on export; and a self-contained player that proves the footage is untouched, even offline.

Starter includes Evidence Lock, so a locked clip cannot be deleted by retention rules. Pro adds the SHA-256 seal at lock time, expiring external share links with multiple recipients, and the self-contained evidence player that proves a clip is untouched in the recipient's browser. Pro also includes full API access and developer SDKs for case-management integrations.

Wherever you choose. Visylix runs on your servers, your virtual machines or your own cloud account, and recordings are written to storage you control, with retention set per camera and per site. Nobody but you holds your footage.

Any identity provider that supports SAML or OpenID Connect, including Okta, Microsoft Entra ID, Google Workspace, Auth0, OneLogin, Keycloak, PingFederate, JumpCloud and Duo, plus Active Directory and LDAP directly.

Privacy redaction is included on Enterprise: faces and plates are found and blurred automatically, and you can draw custom zones for whiteboards, screens or documents caught in frame.

The statute names Huawei, ZTE, Hytera, Hangzhou Hikvision and Dahua Technology, along with their subsidiaries and affiliates. Rebadged units matter: a camera sold under a different brand may still come from a named manufacturer, so check the source, not just the label.

Federal agencies, federal contractors and subcontractors at any tier, and grant recipients using federal funds, plus a growing number of state governments, universities, healthcare systems and critical-infrastructure operators that have adopted the standard in their own procurement rules.

Yes. Fully air-gapped deployment is supported, and it includes the AI. All 22 analytics and the Radha copilot run locally, so no video, metadata or question is sent anywhere.

Yes, through multi-site federation. Sites can be searched and viewed from one operations centre while each keeps its own recordings and retention rules locally. Video stays where it was recorded.

Yes. An operator clicks a remote camera and the browser opens a secure, short-lived live connection directly to that site, so video flows straight from the source site to the operator, with no detour through a central server.

No. Visylix is software. You run it on your own servers, your own virtual machines or your own cloud account, with whichever cameras you choose.

Yes. Evidence locks, seals, audit logs and permissions are copied to the standby continuously, so chain of custody survives the switch, and the audit log records the cutover itself.

Yes. Primary and standby servers can both sit inside your air-gapped network, with no internet connection needed, and licensing works fully offline.

No. Pricing is per plan, not per deployment mode, so on premise, edge, hybrid, air-gapped and your own cloud account all carry the same terms. Starter and Pro include unlimited cameras on the VMS licence; AI analytics are billed per camera per month on the cameras you choose. Plans start at $49 a month, or ₹4,999 a month in India, and Enterprise is quoted per deployment.

Each site is health-checked, and a cross-site search returns every site that answered within a short window, plus a list of any that did not, so one slow link never holds up the result.

High availability pairs a primary server with a standby. If the primary fails, an operator triggers the cutover and the standby takes over in under 60 seconds.

Ask a question such as "find all red cars between 2 and 4 PM yesterday across every site" and Visylix sends it to every site at once. Each site searches its own cameras and returns matches, and Visylix combines them into one feed. Ten sites answer in the time of the slowest one, not the sum of all.

As a Docker image on infrastructure you control: on premise, at the edge, hybrid, fully air-gapped, or in your own cloud account. The same build runs everywhere, so moving between them never means changing product.

Each site has its own signing secret. Every request between sites is signed, tied to the user and the exact operation, and expires within seconds, so a captured request cannot be replayed.

The standby takes over in under 60 seconds after an operator triggers the cutover; the exact time depends on your network's redirection settings. Evidence locks, audit records and permissions are copied to the standby continuously, and up to about 5 seconds of footage per camera may be lost at the moment of cutover. Health checks monitor processor, memory, disk and services.

Running recording and analytics locally at a remote site, such as a depot, branch or plant, with only alerts and key data travelling back to the centre. It suits sites with limited or unreliable bandwidth.

Federation makes many Visylix sites appear as one to the operator. Search every site from one screen, open live view from any site in one click, and receive every site's events in one feed, while video stays where it was recorded. No central archive, no giant network link, no single point of failure.

Yes. Visylix is web-first and runs in any modern browser, including on a phone, so responders can acknowledge alerts one at a time or in bulk from wherever they are. Critical alerts also reach them by SMS, email or webhook, and unacknowledged critical alerts escalate automatically.

Yes. Set up any number of pages, each a named group of cameras with its own layout. The wall cycles through them on an interval you choose and holds on any page with an unacknowledged critical alert until it is acknowledged.

Yes. Every alert carries a severity (critical, high, medium or low). Personal preferences and quiet hours decide which severities reach each person, rules can send critical events to SMS, email, webhook or MQTT, and critical alerts that nobody acknowledges escalate automatically.

Yes. Visylix pushes incidents to Genetec Mission Control, Bosch BIS and any platform that accepts a signed webhook. Each event links straight back to the footage in Visylix.

Yes. Visylix opens recordings straight from Hikvision, Dahua, Uniview, I8H, Bosch and Panasonic recorders, plus CP Plus, Honeywell, Sparsh and Godrej models, so historical footage stays searchable while new recording moves to Visylix. Migration becomes gradual rather than a rip and replace.

Yes. Retention is set per camera and per site, so an entrance can be kept longer than a corridor on the same system, which matters where areas or regions have different legal requirements.

By SMS, email, signed webhooks to Slack, Teams or PagerDuty, MQTT and in-app notifications in any browser, including on a phone. None of them depend on any outside cloud service, and critical alerts that go unacknowledged escalate automatically.

Yes. Pro and above include developer SDKs that cover all 500+ REST APIs: live view control, recordings, AI detections, camera setup, permissions and webhook subscriptions.

Yes. Visylix includes a built-in MQTT broker for sensors and building systems (temperature, smoke, access control, SCADA). Alerts and detections publish to topics you choose, and outside sensors can send events to Visylix so video and sensor data share one timeline.

Yes. Enterprise moves older recordings to your own S3-compatible storage, such as AWS S3, Cloudflare R2, Wasabi, MinIO, Ceph or Cloudian. Archived footage stays searchable and can be brought back on demand.

Yes. Visylix runs on its own native streaming engine, built in-house for surveillance rather than borrowed from a generic media server. It handles 13+ protocols, including NDI and GB28181 ingest, and delivers live view in under a second.

No. Visylix keeps the latest frames ready, so a new viewer starts playing instantly rather than waiting for the camera's next full frame. On cameras with long intervals between full frames, that is the difference between instant video and several seconds of black.

No. AI analytics run on the same video Visylix already handles for live view and recording, on the same servers, with no separate analytics system. Detection zones are set per camera and per analytic, so different analytics can watch different areas of the same view. AI analytics are billed per camera per month on the cameras you choose.

No. Recordings keep the camera's original quality, because Visylix stores the stream without re-encoding it.

Yes. Visylix picks the right stream for each tile automatically. Small tiles in dense grids use the camera's lighter sub-stream, while full-screen and focused views use the main stream. That keeps a full 6×6 wall light on bandwidth and processing.

Connect your access-control system and Visylix compares how many people crossed a door with how many badges were presented there, alerting only when more people entered than credentials. It uses the door-to-camera mapping from your access-control integration and supports Brivo, LenelS2, HID, AMAG and Genea/Openpath, plus others through the generic connector. Without an access-control system, Visylix still detects group crossings from video alone.

Each operator chooses the alert types, cameras and severities they care about. Quiet hours silence non-critical alerts outside working hours, critical alerts go to the on-call team, and acknowledgement tracking prevents duplicate escalation. Sudarshan also scores every alert so noisy sources can be retuned in one click.

Recordings can be sealed with SHA-256 and locked, which stops retention rules deleting them while an investigation is open and gives a chain-of-custody record. Every access and export is written to the tamper-evident audit log, so you can show who viewed or exported a clip and when.

Live view runs over WebRTC and appears in under a second, down to under 200 ms on a good network.

Up to 36 cameras per screen in a 6×6 grid, with layouts from 1×1 to 6×6 chosen per screen. Each browser window runs its own wall, so a control room can open one per display.

Visylix carries large camera counts on standard servers and grows by adding servers and federating sites. We size the hardware with you for your resolution, recording and AI needs.

It depends on camera count, resolution, frame rate, codec and retention rather than any fixed figure. H.265 typically needs about half the storage of H.264 for the same picture quality, so codec choice often matters most. We size storage with you against your own cameras and retention rules.

Yes. White-label branding puts your logo, favicon and branding on every installation, managed through the settings or the API. It is popular with system integrators and managed service providers.

In-app notifications and the alarm console, SMS (Enterprise), email with snapshots, signed webhooks, MQTT, SIEM and PSIM, talkdown through the camera speaker, and siren and door-strike relays. Each channel is chosen per user, and critical alerts that nobody acknowledges escalate.

Signed session tokens for interactive use, scoped API keys for automation, and signed webhooks for outbound events. Every token and key follows the user's permissions and can be revoked at any time.

Visylix protects the picture: a viewer on a weak connection sees slightly less smooth video rather than a frozen or broken image, and recordings are unaffected because they are written from the camera's original stream.

Transmuxing repackages video into a different format without re-encoding the picture. When a viewer can play the camera's own format, Visylix transmuxes rather than transcodes, which keeps original quality and uses very little processing.

Visylix writes the camera's own stream straight to storage without re-encoding it. The recording is an exact copy of what the camera sent, and recording a camera uses almost no processing power, which leaves the server free for live view and AI.

Arrow keys move between tiles, Enter opens the selected tile in focus mode, Escape closes it, and the number keys 1 to 6 switch grid size from 1×1 to 6×6. Operators can run the wall without a mouse.

Visylix runs on its own native engine rather than sitting on top of someone else's media server. Each camera's video is handled once and shared across live view, recording and AI, which is why adding one more camera costs so little.

When the network is unreliable. SRT and RIST recover lost packets and encrypt video over difficult links such as cellular, satellite or long-distance internet, where plain RTSP struggles. RTSP remains the right choice on a clean local network.

No. Visylix is web-first: live view, alerts, investigation and administration all run in any modern browser, on desktop or phone, with nothing to install.

Visylix sends security events directly to Splunk, Datadog and any HTTP collector such as Elastic or QRadar, and every event carries camera, zone, analytic, severity, user and time for correlation rules and SOC playbooks.

Brivo, LenelS2 OnGuard, HID, AMAG Symmetry and Genea/Openpath, plus a generic connector for other systems. Badge reads, door holds and forced-door alarms appear on the same timeline as the video, Sudarshan rules can require a video condition and a badge event together, and rules can switch door-strike relays.

H.264 and H.265 (HEVC) video, and AAC, G.711 and Opus audio. Recordings keep the camera's original stream, so quality is untouched and the server stays free for analytics.

Yes. Visylix runs in any modern browser, including on a phone, with live view over WebRTC, alerts, bookmarks and PTZ control, and nothing to install.

Thirteen plus: WebRTC, RTSP, RTMP, SRT, HLS and Low-Latency HLS, MPEG-DASH, HTTP-FLV, HTTP-TS, NDI, OMT, RIST and GB28181 ingest, plus ONVIF for camera discovery and PTZ control. All run in a native streaming engine built in-house.

Video walls are included on the Enterprise plan, in the browser, with layouts from 1×1 to 6×6, picture-in-picture, auto-cycle and pinning to critical alerts.

No mainstream browser plays RTSP, the format IP cameras use, because browsers work over web protocols that pass through firewalls. Any system showing camera video in a browser converts it first. Visylix converts to WebRTC for sub-second live view, or to HLS where broad compatibility matters more.

Yes. Open your server address in any phone browser, sign in with your Visylix account and you are ready, whether the server is on premise or on an isolated network you can reach.

Yes, through Sudarshan rules. Conditions combine with AND/OR logic and time sequences, so a rule can require several events in order before it fires. Rules can be previewed against past footage first, and actions include alerts, alarm incidents, talkdown, webhooks to terminal systems, and siren relays.

Yes. Beyond the eight industry solutions, Visylix can be shaped for almost any operation. Talk to our team about your requirements and we will configure the right analytics, rules and integrations.

Yes. Unique person counting counts each visitor once per window, which removes double counting from footfall. That figure can be joined with your point-of-sale data for conversion rates.

Yes. Detection zones are set per camera and per analytic, so a press area can require PPE while a walkway does not. Sudarshan rules combine conditions and time sequences and act on violations with alerts, talkdown through the camera speaker and siren relays. Machine interlocks stay under your own control systems.

Yes. Visylix measures classroom and facility occupancy, parking availability and foot-traffic heat maps, helping administrators plan rooms, maintenance and security patrols.

Yes. Visylix provides queue and wait-time analytics with predicted waits, footfall counting, and watchlist alerts for VIPs and known offenders, helping branch managers understand waits and staffing. 500+ REST APIs connect it to banking systems.

Yes. Visylix connects to existing cameras through ONVIF and RTSP, ingests GB28181 devices, opens recordings on existing NVRs, and offers 500+ REST APIs for city management systems and command centres.

Yes. Visylix offers 500+ REST APIs, MQTT and 13+ streaming protocols for integration with traffic management, toll and fleet systems.

Yes. 500+ REST APIs connect Visylix to hospital information, nurse call and visitor management systems, and it works with the cameras the hospital already has.

Yes. Manage every site from one platform with multi-site federation, large camera counts on each server, 500+ REST APIs, and archive access to existing Hikvision, Dahua, Uniview, CP Plus and other recorders.

Yes. Queue and wait-time analytics and heat maps are two of the 22 AI analytics, running on the same cameras you already record. Detection zones are set per camera and per analytic, so a till area and an aisle can be measured differently on the same view.

Yes. Visylix watches production areas with anomaly detection, PPE, intrusion zones and line-crossing counts, catches the unusual event nobody wrote a rule for, and alerts in real time through the console, email, webhooks or MQTT.

Yes. Visylix runs on the bank's own infrastructure, including fully air-gapped. All 22 AI analytics and the Radha copilot run on premise, so branch footage and every question about it stay inside the bank network.

Yes. Visylix runs fully air-gapped. All 22 AI analytics and the Radha copilot run on premise, so no video, data or question is sent outside.

Yes, by number plate. Plate journeys are exact: the same plate read at different cameras becomes one timeline with travel times between reads. Person journeys follow someone by appearance to trace a route through a terminal quickly.

Yes. 46 permissions across 4 roles mean a head teacher, a site manager and a security officer can each see different cameras and functions. Every view and export is audit-logged, so you can show who accessed what footage and when.

Yes, through multi-site federation. Each store records locally while a central team watches every site, with permissions deciding who sees which store and retention rules set per site.

Yes. Removed-object detection runs alongside abandoned-object detection, so a tool or part taken from a marked area raises an alert. Anomaly detection, which learns each camera's normal, covers the unusual case nobody wrote a rule for.

Yes. Sudarshan rules combine conditions and time sequences, so a rule can require several things to happen in order before it fires, and can be previewed against past footage before going live. Actions include alerts, alarm incidents, talkdown, siren relays and escalation.

Yes. Sudarshan rules can require both a video condition and an access-control event, so a door opening without a matching badge raises an alert while an ordinary authorised entry does not. Conditions can be ordered into sequences and previewed against past footage.

Not on the VMS licence. Starter and Pro include unlimited cameras, so adding stores or cameras does not change the plan price. AI analytics are a separate subscription, billed per camera per month only on the cameras where you want them. Plans start at $49 a month, or ₹4,999 a month in India.

Visylix is engineered for the frameworks our buyers are audited against. It runs on your own infrastructure and provides a tamper-evident audit trail, retention rules per camera, redaction on export, data export and erasure, permission-level access control and sealed evidence, the controls auditors look for in healthcare, banking and government.

Yes. Visylix provides evidence locks sealed with SHA-256, retention set per camera and per site, and a tamper-evident audit trail of every video access. It runs fully on premise, including air-gapped.

Visylix runs fully on premise, including air-gapped, with 46 permissions across 4 roles, retention rules, redaction on export, data export and erasure, and a tamper-evident audit trail. It is engineered for the frameworks our buyers are audited against.

Yes. Visylix runs fully on the hospital's own infrastructure, so patient video never leaves the building, including AI analysis and copilot questions. It adds 46 permissions across 4 roles, retention rules, redaction on export and a tamper-evident audit trail, and it is engineered for the frameworks our buyers are audited against.

Yes. Radha produces shift briefings and printable incident reports, and Visylix keeps a full audit trail of every PPE and zone alert for safety audits and insurers. Everything stays on premise, including air-gapped plants.

Yes. Visylix provides number plate recognition with allow, block and watch lists, exact plate journeys and vehicle counts across ports, warehouses and distribution centres, and opens historical footage on your existing recorders.

Yes. Object detection, intrusion detection for restricted areas, line crossing, watchlist alerts and talkdown work together for loss prevention, with real-time alerts through the console, email, webhooks or MQTT.

It is available, but it is a policy decision, not a default. Many education authorities restrict biometric processing of minors, so most schools use intrusion, loitering, line crossing, crowd, weapon and audio alerts instead, which answer the safety questions without processing faces.

Yes. Visylix runs with no internet connection at all. All 22 AI analytics and the Radha copilot run locally, so plant footage and every question about it stay inside the plant network.

No. Multi-site federation lets several deployments across districts be monitored from a central command centre, with permissions and retention set per site. Each site keeps its recordings locally, which suits separate data-custody rules.

Radha runs fully on the hospital's own hardware, so nothing it handles leaves the building. Its 111 tools follow each user's permissions, its 13 high-impact actions always ask for confirmation, it has no delete tool, and every question and action is audit-logged.

No. Visylix runs on infrastructure the city controls, including fully air-gapped. All 22 AI analytics and the Radha copilot run locally, so no video or data goes to an outside service.

PPE detection scores mAP@50 0.865 on a frozen 10,315-image evaluation set.

Visylix provides perimeter intrusion detection with after-hours alerts, weapon and audio alerts, floor-plan maps with live camera status, watchlist alerts, talkdown, and sealed evidence sharing with police. It carries large camera counts on each server across multi-building campuses.

Visylix detects patient falls around the clock with instant alerts to nursing staff, watches restricted areas such as pharmacies, and measures emergency department occupancy. Everything runs on the hospital's own servers, so patient video stays inside the building.

SRT and RIST are built for exactly this, recovering lost packets and encrypting video over links that drop. Edge deployment is supported too, so recording and analytics run locally at a depot or station and only alerts travel back to the centre.

Visylix builds cross-camera journeys two ways. Number plate journeys are exact: the same plate read at many cameras is stitched into one timeline with travel times between points. Person journeys follow someone by appearance across cameras, which is fast for tracing a route, with an operator confirming each sighting.

Visylix counts each visitor once for true footfall, maps shopper journeys, shows heat maps for merchandise placement, measures queues and predicts waits, and alerts staff when a known repeat offender walks in.

Loitering, intrusion, tailgating, abandoned object and camera tamper detection all run on the cameras you already record, with zones set per camera and per analytic. Tamper detection learns each camera's view, which matters at unattended sites where nobody would notice a moved or covered camera.

Visylix checks helmets and vests by default and up to nine items of protective equipment, including goggles and gloves, with zone-by-zone rules around heavy machinery. Each camera learns its own normal, and when an operator marks a PPE alert as a false alarm, that camera tightens automatically, never below the operator's own setting.

Visylix provides number plate recognition with real-time watchlists and exact plate journeys, anomaly, stopped-object and intrusion alerts on highways and in tunnels, speeding, wrong-way and U-turn alerts, and passenger flow for terminals and stations, all on your own infrastructure.

Through 500+ REST APIs, signed webhooks, MQTT, SIEM and PSIM. Access-control events can be combined with video, so an alert can require both a badge event and a camera detection. Single sign-on, Active Directory, LDAP and passkeys handle operator sign-in.

By plan. Starter and Pro include unlimited cameras on the VMS licence, so adding cameras or sites does not change the plan price, and AI analytics are billed per camera per month only on the cameras you choose. Plans start at $49 a month, and Enterprise is quoted per deployment.

Visylix detects loitering, intrusion, camera tampering and abandoned objects at ATMs in real time. In branches, watchlist alerts flag known offenders and VIP customers. Tamper detection learns each camera's view, and every alert carries a dispatch score so noisy sources can be retuned.

Visylix combines face watchlists checked against badge events, perimeter intrusion with virtual perimeters, after-hours alerts, PTZ autotracking, presets and patrol tours, siren and door-strike relays, and a tamper-evident audit trail. Each camera learns its own normal, and marking a false alarm tightens it automatically, never below the operator's own setting.

Through on-premise processing and redaction. Visylix runs on hospital infrastructure, so no footage leaves the building, including AI analysis and copilot questions. Faces and custom zones are redacted on export, and 46 permissions across 4 roles control who sees which cameras.

Visylix adapts to the store. Anomaly detection learns what normal looks like for each camera, and tamper detection learns each view. When an operator marks an alert as a false alarm, that camera's threshold for PPE, object, fire and weapon detection tightens automatically, never below the operator's own setting. Every alert also carries a dispatch score, and the effectiveness board proposes retunes an operator approves in one click.

Visylix runs city-wide camera networks from one platform: traffic and number plates, crowd density and stampede alerts, public safety, abandoned objects, weapons and fire, with Sudarshan rules that act on their own and sealed evidence for courts. It carries large camera counts on each server with sub-second live view, and federation joins districts into one command centre.

Per plan. Starter and Pro include unlimited cameras on the VMS licence, so a plant with 600 cameras carries the same plan price as one with 60. AI analytics are billed per camera per month on the cameras you choose, and plants can be federated into a central view while each records locally. Enterprise is quoted per deployment.

Per plan. Starter and Pro include unlimited cameras on the VMS licence, so adding cameras or branches does not change the plan price. AI analytics are billed per camera per month on the cameras you choose. Plans start at $49 a month, and branches can be federated into a central view while each keeps recordings locally.

As long as you need. Retention is set per site and per camera, so different wards can keep footage for different periods, and evidence lock holds specific recordings while an investigation is open.

Retention is set per camera and per site, so entrances can be kept longer than corridors if policy requires. Evidence lock holds specific recordings while an incident is investigated, so routine deletion never removes material that is still needed.

Visylix carries large camera counts on standard servers and grows by adding servers, with multi-site federation joining sites into one estate.

Typically within hours. Visylix installs as a Docker deployment that tunes itself to your hardware, and our team configures the right analytics and rules for your sector.

Visylix gives auditors what they look for: fully on-premise or air-gapped deployment, encrypted connections, 46 permissions across 4 roles, retention rules, redaction on export and a tamper-evident audit trail of every operator action. It is engineered for the frameworks our buyers are audited against.

Visylix runs on premise with encrypted connections, 46 permissions across 4 roles, retention rules, evidence locks sealed with SHA-256 and a tamper-evident audit trail of every action. It is engineered for the frameworks our buyers are audited against.

Yes. Visylix suits everything from a single school to a large university. It runs fully on premise, controls access with 46 permissions across 4 roles, and offers 500+ REST APIs for school management systems.

Yes. Visylix is built in Kolkata, India by Aptibit Technologies, with an interface in 55 languages including 13 Indian languages and a copilot that answers in Hindi and Hinglish. It runs fully air-gapped for government sites, ingests existing estates through ONVIF, RTSP and NVR archive access including CP Plus, Honeywell, Sparsh and Godrej, and is built for Safe City and Smart City command centres.

Yes. Radha runs entirely on premise with 111 tools, every one limited by the user's own permissions. Its 13 high-impact actions always ask for confirmation, it has no delete tool, and every action is written to the audit log with the operator who asked for it.

Starter and Pro include unlimited cameras on the VMS licence, so a terminal with 400 cameras carries the same plan price as one with 40. AI analytics are billed per camera per month on the cameras you choose. Plans start at $49 a month and Enterprise is quoted per deployment. Visylix runs on your own infrastructure, so there is no hardware to buy from us.

The VMS licence covers unlimited cameras on Starter and Pro, so it does not grow with camera count. AI analytics are a separate subscription, billed per camera per month on the cameras you choose; talk to us for AI pricing. Plans start at $49 a month, or ₹4,999 a month in India, and Enterprise is quoted per deployment.

Starter and Pro include unlimited cameras on the VMS licence, from $49 a month or ₹4,999 a month in India. AI analytics are billed per camera per month on the cameras you choose. A district can federate many schools into one central view while each keeps its own recordings and retention rules.

They usually stay. Visylix discovers ONVIF cameras automatically and connects over RTSP, so most IP cameras from the last decade work without replacement. Recordings on existing Hikvision, Dahua, Uniview, CP Plus and other recorders stay reachable while new recording moves to Visylix.

Visylix serves smart cities and police, enterprise security, retail, transportation, healthcare, manufacturing, education, and banking and finance, with analytics, rules and workflows shaped for each.

Rules vary by country and state, but cameras are generally acceptable in common areas such as corridors, entrances, grounds and car parks, and not in bathrooms, changing rooms or other private spaces. Many jurisdictions also require signage and a documented retention period. Check your local education authority rules; this is general guidance, not legal advice.

Visylix supports 13+ natively. RTSP and ONVIF cover fixed IP cameras, SRT and RIST carry video reliably over cellular or long-distance links, GB28181 brings in devices built to that standard, and WebRTC delivers sub-second live view to control-room operators.

In most cases, yes. Visylix discovers ONVIF cameras and connects to RTSP streams directly, so it works with the cameras you have. Existing NVRs stay reachable for historical footage, so a rollout never requires replacing recorders on day one.

Yes. Visylix works with ONVIF cameras and any RTSP source, so your existing cameras carry over unchanged. Our team helps move recording settings, user permissions and layouts.

Yes. Visylix works with ONVIF cameras and any RTSP source, so your existing cameras carry over unchanged, and our team helps move recording settings and permissions. Visylix also connects natively to Brivo, LenelS2, HID, AMAG and Genea/Openpath and combines badge events with video in Sudarshan rules.

If your cameras support ONVIF, yes. Visylix connects to ONVIF IP cameras from any maker. Rhombus cameras are built for Rhombus's own cloud, so those units would be replaced with ONVIF cameras over time.

Yes. Visylix works with ONVIF cameras from virtually any maker. Keep your existing Axis, Hikvision, Dahua, Hanwha, Bosch, CP Plus or other cameras and add Visylix as your video management and AI platform.

Yes. Visylix runs fully air-gapped. Video management, all 22 AI analytics and the Radha copilot with 111 tools work completely offline, and because the AI and the copilot are built in-house, there is no third-party plugin to license or connect.

Yes. Visylix runs on premise, at the edge, hybrid, fully air-gapped or in your own cloud account. All video, AI analysis and the Radha copilot with 111 tools stay inside your infrastructure.

Yes. Video management, recording, live view, AI analysis and the Radha copilot all work with no internet connection.

Yes. Visylix runs fully air-gapped, with video management, AI analysis, recording and the Radha copilot all working offline.

No. Visylix records natively, with continuous and event recording, storage management and playback, so no separate NVR is needed. It can also open recordings on the NVRs you already own.

Visylix offers 22 AI analytics built in-house, from face and number plate recognition to weapons, fire, PPE, crowds and audio events, running on cameras you already own. Each camera learns its own normal, and false alarms tighten thresholds automatically. Visylix also includes Radha, an on-premise AI copilot with 111 tools that answers in English, Hindi and Hinglish.

Yes. Visylix connects natively to Brivo, LenelS2, HID, AMAG and Genea/Openpath, plus others through a generic connector. Badge events join the video timeline, Sudarshan rules combine them with detections, and rules can switch door-strike relays. You keep your choice of access-control system rather than being locked to one vendor.

No. Visylix records natively, so no separate NVR is needed, and it can still open recordings on the NVRs you already own.

Yes. Visylix runs entirely within your infrastructure, including the Radha copilot, on premise, at the edge, hybrid, air-gapped or in your own cloud account. It also opens recordings on existing NVRs and offers 500+ REST APIs for deep integration.

Yes. Visylix works with ONVIF cameras from the same broad range of makers, adds device packs for the top five camera brands, and accepts any camera that serves RTSP.

Yes. Visylix works with ONVIF cameras from Axis, Hikvision, Dahua, Hanwha, Bosch, CP Plus and hundreds of other makers. You are never locked into one camera vendor.

Visylix offers 22 AI analytics built in-house: face recognition, number plates, object detection, pose, crowds, PPE, heat maps, motion, person tracking, unique person counting, intrusion, line crossing, camera tampering, abandoned objects, tailgating, queues, parking, speed, anomalies, fire and smoke, weapons and audio events. Each camera learns its own normal and false alarms tighten it automatically, and Sudarshan rules turn detections into action.

Visylix offers 22 AI analytics built in-house, covering people, vehicles, safety and sound, and runs them on the cameras you already own. Each camera learns its own normal, and false alarms tighten thresholds automatically. Visylix also includes Radha, an on-premise AI copilot with 111 tools.

Visylix runs entirely on your infrastructure: all video, AI analysis and the Radha copilot with 111 tools stay inside your network. Nothing leaves your building, which suits strict data-sovereignty requirements.

Visylix handles large camera counts on each server across 13+ protocols, and multi-site federation joins sites into one estate with central management and site-level control, so you can run thousands of cameras across many locations from one interface.

Visylix offers 22 AI analytics built in-house, so there are no plugins to buy, install or maintain. Each camera learns its own normal, false alarms tighten thresholds automatically, and Sudarshan rules turn detections into action.

Visylix runs on your own infrastructure as a Docker image, on premise, at the edge, hybrid, air-gapped or in your own cloud account. All video, AI analysis and the Radha copilot with 111 tools stay inside your network.

Visylix plans start at $49 a month, and Starter and Pro include unlimited cameras on the VMS licence, so the plan price does not rise as you add cameras. AI analytics are billed per camera per month only on the cameras where you want them.

Visylix goes beyond search with Radha, an on-premise AI copilot with 111 tools that can search recordings, control cameras, manage alerts, configure analytics and write automations in plain language.

Visylix is a complete video management system with native recording, live view, camera management and 22 AI analytics built in-house, plus Sudarshan automation and the Radha copilot. It replaces your VMS stack rather than adding a layer on top of it.

Visylix plans start at $49 a month, and Starter and Pro include unlimited cameras on the VMS licence, so you can grow to thousands of cameras without the plan price changing. AI analytics are billed per camera per month only where you want them.

Yes. Visylix publishes its prices: Starter at $49 a month and Pro at $99 a month, both with unlimited cameras on the VMS licence. AI analytics are billed per camera per month on the cameras you choose, and Enterprise is quoted directly.

Yes. Visylix plans start at $49 a month with unlimited cameras on the VMS licence, and you can use the cameras you already own. AI analytics are billed per camera per month only on the cameras where you want them.

Yes. Visylix is built for enterprise scale, with large camera counts on each server, 13+ protocols and a native streaming engine built in-house, plus federation for multi-site estates.

Verkada uses its own cameras, so the move is gradual: add ONVIF cameras alongside existing units and run Visylix in parallel. Our team helps plan a phased migration.

Visylix is a complete video management system with native recording, live view, camera management, 22 AI analytics built in-house, Sudarshan automation and the Radha copilot. It replaces your video management stack rather than layering analytics on top.

Visylix offers 22 AI analytics built in-house, covering people, vehicles, safety and sound, with no partner plugins required. Each camera learns its own normal, false alarms tighten thresholds automatically, and Sudarshan rules turn detections into action.

Visylix handles large camera counts on each server across 13+ protocols, runs on premise, opens recordings on existing NVRs, offers 500+ REST APIs and federates sites, which makes it a strong fit for large, distributed estates.

Visylix is built for enterprise scale, with large camera counts on each server, 13+ protocols, a native streaming engine and multi-site federation.

Yes. Visylix runs entirely on your own infrastructure and works fully air-gapped after installation. All video, AI data, face watchlists and Radha conversations stay within your network. It also runs at the edge, hybrid or in your own cloud account.

Yes. Visylix runs in completely disconnected environments. All AI analytics, live view, recording and the Radha copilot run locally, which suits government, defence and classified networks.

Yes. Visylix converts H.265 video for browsers that cannot play it, and converts camera audio for browser playback, so operators never have to think about formats.

Yes. Visylix sends events directly to Splunk, Datadog and any HTTP collector such as Elastic or QRadar, and pushes incidents to Genetec Mission Control, Bosch BIS and any platform that accepts a signed webhook, with a link back to the footage.

Yes. Alongside number plate reads, Visylix recognises vehicle types such as car, truck, bus and motorcycle and tracks direction of travel, so every vehicle can be searched by plate and by type.

No. Visylix runs entirely on your own infrastructure. All video, AI analysis, recording and storage stay on your servers, and nothing leaves your network unless you choose to connect it.

Yes. Enterprise includes high availability with primary and standby servers and an operator-initiated cutover in under 60 seconds, and multi-site federation lets one operator search, receive events from and watch cameras across every site while video stays where it was recorded.

Yes. Visylix opens recordings straight from Hikvision, Dahua, Uniview, I8H, Bosch and Panasonic recorders, plus CP Plus, Honeywell, Sparsh and Godrej models. Search by time range, play several channels side by side at 0.5× to 8×, and download footage sealed with SHA-256.

Yes. Fisheye cameras can be panned and zoomed inside the 360° image in live view (Enterprise). Enterprise also includes browser video walls up to 6×6, with auto-cycle and pinning to critical alerts, one wall per display.

Start a free 7-day trial, or contact our team at info@aptibit.com for a personalised demo. We will understand your requirements, size the hardware with you and plan the rollout. A standard install is a single Docker deployment on your own server.

Evidence lock seals each recording with SHA-256 at lock time and exports a self-contained player that checks the clip offline. The audit trail logs every view, export and share with time, user and source. Case sharing (Pro and above) supports expiring signed links and view limits, and erasure requests can be honoured while keeping the evidence seal intact.

Visylix protects the picture under pressure: viewers on weak connections see slightly less smooth video rather than a frozen image, the server stays responsive at peak load instead of failing, and new viewers start playing instantly.

Visylix keeps all personal data in your chosen jurisdiction by running on your own infrastructure, and adds retention rules per camera, redaction on export, data export and erasure, consent records for face enrolment and a tamper-evident audit trail.

Alerts reach people through in-app notifications and the alarm console, SMS (Enterprise), email with snapshots, signed webhooks and MQTT. Personal preferences, quiet hours, severity routing, acknowledgement tracking and automatic escalation make sure the right person responds.

Visylix records continuously, on a schedule, on events (with time before and after) or manually. It stores the camera's own stream without re-encoding, so quality is untouched and recording uses almost no processing, and H.265 is recorded natively for smaller storage bills. Recordings include thumbnails, retention set per camera, and timeline playback up to 8×.

Visylix offers 22 AI analytics, all built in-house. People: face recognition with watchlists and unknown-person alerts, person tracking, unique person counting, crowd detection with density and stampede alerts, fall and running detection, tailgating, queues and heat maps. Vehicles: number plate recognition (ANPR) with allow, block and watch lists and exact plate journeys, parking occupancy, and speeding, wrong-way and U-turn alerts. Safety and security: intrusion, line crossing, weapons, fire and smoke, PPE, abandoned and removed objects, and camera tampering. Scene and sound: object detection across 80 classes, motion, anomaly detection and audio events. Each camera learns its own normal, and marking a false alarm tightens it automatically, never below the limit you set. AI analytics are billed per camera per month on the cameras you choose.

Starter and Pro cover unlimited cameras on the VMS licence, so adding cameras never changes the plan price. Each server carries large camera counts, and multi-site federation joins sites into one estate. We size the hardware with you, and the installer tunes itself to the server it runs on.

Visylix carries large camera counts on standard servers and grows by adding servers, with multi-site federation for the largest estates.

Yes. Visylix runs on your own infrastructure and provides the tools privacy rules ask for: retention rules per camera, redaction on export, data export and erasure, consent records for face enrolment, encrypted credentials and a tamper-evident audit trail. Its interface is available in 55 languages. Visylix is engineered for the frameworks our buyers are audited against.

Visylix starts with a free 7-day trial on 1 camera. Starter is $49 a month ($539 a year) and Pro is $99 a month ($1,089 a year). In India, Starter is ₹4,999 a month (₹54,989 a year) and Pro is ₹9,999 a month (₹1,09,989 a year), plus GST. Both cover unlimited cameras on the VMS licence. AI analytics are a separate subscription, billed per camera per month on the cameras you choose; talk to us for AI pricing. Enterprise is priced to your estate.

Any of the 22 AI analytics built in-house, chosen per camera: face recognition, number plates, PPE, intrusion, tailgating, queues, parking, fire and smoke, weapons, audio events, anomaly detection and more. AI analytics are a separate subscription, billed per camera per month on the cameras you choose.

Enterprise includes single sign-on through SAML and OIDC (Okta, Microsoft Entra ID, Google Workspace, Auth0, OneLogin, PingIdentity and others) plus Active Directory and LDAP. Two-factor sign-in with an authenticator app and recovery codes is available from Starter, and passkeys from Pro, compatible with Apple and Google passkeys, password managers and hardware keys.

Brivo, LenelS2 OnGuard, HID, AMAG Symmetry and Genea/Openpath, plus a generic connector. Badge reads, door holds and forced-door alarms appear on the same timeline as the video, Sudarshan rules can require a badge event and a detection together, and rules can switch door-strike relays.

Enterprise moves older recordings to your own S3-compatible storage, such as AWS S3, Cloudflare R2, Wasabi, MinIO, Ceph or Cloudian. Archived footage stays searchable and can be brought back on demand.

Unlimited detection zones and directional tripwires per camera. Zones decide which analytics run where, with their own thresholds and schedules, and tripwires count entries and exits separately.

Visylix encrypts every connection, encrypts live video in the browser and encrypts stored credentials. Recordings sit on storage you control and can encrypt, and encryption keys stay on your own infrastructure.

Radha is the AI copilot built into Visylix, and the only VMS copilot we know of that holds a conversation, takes action and runs fully on premise. It runs on an on-premise language model on your own hardware and has 111 tools: control of cameras and recordings, answers from every AI analytic, video search by description, multi-step investigations with evidence timelines, live scene description, voice input, daily briefings, incident reports and administration. It answers in English, Hindi and Hinglish, remembers your site's vocabulary, confirms every high-impact action and logs everything, all fully air-gapped.

Visylix is automated AI video surveillance, built in India by Aptibit Technologies. It connects to your existing cameras over RTSP, ONVIF and 13+ protocols, watches every feed with 22 AI analytics, decides what matters with Sudarshan rules, and acts with alerts, sirens, door-strike relays, talkdown and escalation. Radha, the AI copilot, runs it all in plain language. It also opens recordings on existing NVRs, offers 500+ REST APIs, and runs fully on premise or air-gapped.

Visylix is automated AI video surveillance, built in India by Aptibit Technologies. Instead of recording footage for someone to review later, it watches every camera with 22 AI analytics, decides what matters with Sudarshan rules, and acts on its own with alerts, sirens, door-strike relays, talkdown and escalation. Radha, the built-in AI copilot, runs the whole system in plain language. It is a complete video management system with video walls, maps, an alarm console, PTZ autotracking, sealed evidence export and sub-second live view, and it replaces traditional CCTV, DVR and NVR software.

The Visylix interface is available in 55 languages, including 13 Indian languages (Hindi, Bengali, Tamil, Telugu, Kannada, Malayalam, Assamese, Marathi, Gujarati, Punjabi, Rajasthani, Haryanvi and Odia), four right-to-left scripts (Arabic, Hebrew, Persian and Urdu), every major European language, and Chinese, Japanese, Korean, Thai, Vietnamese, Indonesian, Malay and more. It detects the browser language automatically and each user can choose their own. Radha, the copilot, answers in English, Hindi and Hinglish.

Redaction is applied before footage is released to anyone. Enterprise includes full redaction of faces, plates and custom zones on exports and shared clips, and authorised roles such as investigators can view unredacted footage, with every view audit-logged.

Radha checks every request before acting, follows each user's permissions on all 111 tools, limits how fast any one user can send requests, and logs every action. Attempts to trick it with hidden instructions are blocked and recorded, and its 13 high-impact actions always ask for confirmation.

Visylix supports 13+ protocols in a native streaming engine built in-house: WebRTC for sub-second live view, down to under 200 ms; RTSP; RTMP; SRT; HLS and Low-Latency HLS; MPEG-DASH; HTTP-FLV; HTTP-TS; NDI; OMT; RIST and GB28181 ingest; and ONVIF for camera discovery and control.

The Visylix blog covers AI video analytics, streaming protocols (WebRTC, RTSP, HLS, SRT), edge surveillance, deployment, industry solutions (smart cities, retail, manufacturing), comparisons, migration guides and video security best practice.

Articles are written by Falguni Chatterjee, Founder and CEO of Aptibit Technologies, and the Visylix engineering team, drawing on the experience of building an AI video surveillance platform from the ground up.