Zaylix vs Lumana
If you're evaluating Lumana for AI video security, here's a factual comparison to help you decide what fits your situation. We're not going to pretend to be a bigger or more established company than we are, Zaylix is early-stage; Lumana is a funded, more established platform.
Lumana
Lumana is an AI video security platform built around its own hardware ecosystem (cameras, on-prem appliances like Core and VIA-1) alongside cloud AI agents for monitoring, response, investigation and insight. It's positioned for mid-size to enterprise deployments and sells both hardware and software.
Zaylix
Zaylix is a software-only AI layer that runs on the CCTV and NVR infrastructure you already have. We don't sell cameras or appliances. We're an early-stage company: our detection models are tested and working, but we don't yet have a completed, independently-owned live client deployment, which we tell every prospect upfront.
| Point | Lumana | Zaylix |
|---|---|---|
| Hardware | Sells its own camera and appliance hardware as part of the platform. | Runs on cameras and NVRs you already own. No hardware purchase required. |
| Company stage | Established, funded platform with existing enterprise deployments. | Early-stage. Tested detection models and POC footage, no completed live deployment yet, disclosed honestly. |
| Biometrics | Not something we've independently verified; check their current documentation. | Biometric-free by design, no facial recognition, built this way from the start. |
| Before you buy | Standard demo and sales process. | Written feasibility review first: what will and won't work on your specific cameras, before any commitment. |
Who each is a better fit for
If you want an established, funded vendor with its own hardware ecosystem and a longer track record, Lumana is worth evaluating directly. If you want to add AI detection to cameras you already have, without buying new cameras, and you're comfortable working with an early-stage team that's upfront about where it is, Zaylix may be a fit.
How Zaylix actually works
Zaylix connects to your existing cameras over RTSP or ONVIF, the standard streaming protocols almost every IP camera and NVR already supports, so there's typically no change to your camera setup. The video stream is analyzed on dedicated inference infrastructure, where detection models run continuously, looking for the specific events each site is configured to flag: a person entering a restricted zone, missing PPE, smoke, an unrecognized plate, and so on. When a model flags something, it doesn't go straight to the customer. A human reviewer checks the clip first, the step that cuts down false alarms, before an alert goes out by whichever channel the customer prefers (app, SMS, call, or a webhook into their own systems). The whole loop, detection to verified alert, is designed to run in real time, not as an end-of-day report.
What your cameras need
- An IP camera or NVR that streams over RTSP or ONVIF, which covers the large majority of commercial CCTV installed in the last 10-15 years.
- Resolution and mounting angle that actually captures the thing you want detected, a wide shot of a parking lot won't reliably catch a missing hard hat. We check this per camera in the feasibility review, not after a contract is signed.
- A stable network path from the camera (or NVR) to wherever the inference runs, on-prem or cloud depending on the deployment.
- Adequate lighting for the detection type in question; low-light PPE or ANPR detection needs either IR-capable cameras or supplemental lighting, which we'll flag if it's a problem.
What Zaylix can't do
- We don't manufacture or sell cameras, NVRs, or any physical hardware. If a site's existing cameras genuinely can't support a detection type (wrong angle, too low resolution, no night vision where it's needed), we say so in the feasibility review rather than force a sale.
- We're early-stage: there's no completed, independently-owned live client deployment to point to yet, only tested models and POC footage. We tell every prospect this upfront, and the free feasibility audit and pilot exist specifically so you can verify fit before committing.
- Human verification adds a short delay between a detection and the alert reaching you, it's not instantaneous AI-only alerting. That's a deliberate tradeoff to cut false positives, not a bug, but worth knowing if you need sub-second automated response.
- We don't do facial recognition or any biometric identification, by design. If a use case specifically requires identifying who a person is (not just that a zone was entered), Zaylix isn't built for that.
Frequently asked questions
Does switching from Lumana (or to Zaylix) mean replacing our cameras?
No, in either case the pitch from both companies is that AI runs on top of existing or newly supplied cameras. The real difference is that Lumana sells its own camera and appliance hardware as part of the platform, while Zaylix is software-only and works with whatever CCTV or NVR infrastructure you already have.
Is Zaylix a drop-in replacement for Lumana?
Not necessarily, and we wouldn't claim that. Lumana is a funded, more established platform built for mid-size to enterprise deployments with its own hardware ecosystem. Zaylix fits a narrower, earlier-stage use case: businesses that want AI added to cameras they already own, and are comfortable working with an early-stage vendor that's upfront about where it is.
What detection types does Zaylix support?
PPE and safety compliance, intrusion and perimeter detection, fire and smoke detection, ANPR (license plate recognition), occupancy and loitering detection, and driver drowsiness and distraction monitoring. Each is detailed on its own solutions page.
How does the human verification step work?
When a detection model flags an event, a human reviewer checks the clip before any alert reaches the customer. This is what cuts down false positives from things like shadows, steam, or reflections being misread as something real.
Do you use facial recognition?
No. Zaylix is biometric-free by design, we detect behavior and presence in a zone, not identity.
What does the feasibility review actually check?
Camera resolution, mounting angle, lighting conditions, and network connectivity for each camera you want covered, matched against what the specific detection type needs. It happens before any commitment, so you know what will and won't work on your actual site.
Is there a cost to evaluate Zaylix?
No. The feasibility audit and the initial pilot are free. We only discuss pricing once you've seen it working on your own footage.