Recorded, not seen.
CCTV runs 24/7 across every site. Humans review almost none of it. The recording light does not translate to a pair of eyes—and incidents happen in the gap between what was filmed and what was noticed.
Closed by Monitor
Monitor / Search / Prove
Every camera, a trained pair of eyes.
SIMIS turns passive CCTV into a video intelligence platform. Proprietary detection, searchable archive, verifiable records—running locally on the cameras you already own.
Multi-model AI stack. No facial recognition. No cloud dependency.
AI Singapore 1st Place, Pan-SEA Developer Challenge 2025 SUTD ARISE Validate, 2026
The gap
Your cameras record everything. You can use almost none of it.
Cameras are everywhere. Awareness, search and chain of custody are not. Three gaps in the status quo; each one is what SIMIS is built to close.
Archived, not searchable.
When something does happen, finding the moment means dragging a timeline by hand, frame by frame. The standard tool is a scrub bar: manual, linear, and far slower than the question deserves.
Closed by Search
Forensic by accident.
When you do find the right clip, the chain of custody is whatever you can later reconstruct from logs, memory and a grainy file off a hard drive. The moments you most need to defend are the ones the system was least designed for.
Closed by Prove
The platform
Three layers, from live alert to verifiable proof.
SIMIS is not a smarter alert. It is the link between real-time monitoring, searchable history and the record you can show—built so each layer feeds the next. Starting with CCTV; the same engine extends to any video stream you operate.
01 Today
Monitor
Proprietary detection, in real time.
A multi-model AI stack reads movement, posture and proximity from any RTSP or ONVIF camera you already own—pose-based, not face-based, so privacy is preserved by construction. Alerts surface the moment something looks wrong.
- Multi-model AI stack
- Local inference
- No faces captured
02 In development
Search
Find footage by what happened, not when.
Flagged events are indexed with vector embeddings so you can retrieve them in plain language. "Two children near the back gate, after 3pm" returns clips, not timestamps to scrub through.
- Natural-language queries
- Visual embeddings
- Seconds, not hours
03 On roadmap
Prove
Show what happened, when someone asks.
Every detection writes to a time-stamped, cryptographically linked record. When a parent, regulator or insurer asks what happened, you can show them—not just describe.
- Time-stamped log
- Verifiable record
- Export ready
Why SIMIS
Four decisions you cannot bolt on later.
Privacy by architecture
We capture pose, not faces. The data we store cannot be re-identified to individuals—PDPA and GDPR friendly by construction, not by promise.
Search, don’t scrub
Reviewing CCTV today means dragging a timeline. SIMIS turns the archive into a question you can ask in English and get answered in seconds.
Local processing
Inference runs on a small box on site. No cloud round-trips, no streaming footage off-premises, no recurring egress bills.
Sold as a system
Hardware, a software subscription, and on-site setup and training—from one team. Plugs into the cameras you already own; we own the rest. One number to call when something needs to change.
Endorsement
Backed by AI Singapore.
AI Singapore is the country’s national AI programme, hosted by the National University of Singapore and funded by the National Research Foundation. They named SIMIS the 1st place winner of the Pan-SEA Developer Challenge, Healthcare Track, 2025.
The winning pitch, AI Singapore Pan-SEA Developer Challenge final, 2025.
Since then
- 2026
Accepted, SUTD ARISE
Selected for the SUTD ARISE Validate programme—early-stage deep-tech support and validation runway.
- April 2026
Incorporated in Singapore
Founded with healthcare-grade engineering experience across the team. Pre-seed.
- Now
Pilot conversations
Active discussions with childcare operators in Singapore. Pilot scope, hardware spec and success criteria being defined.
Team
Healthcare people who build AI.
Three founders. One has shipped regulated healthcare. One has shipped production AI inside government and pharma. One has shipped hardware out of Shenzhen. We have seen what it takes to put serious software in front of serious users.
J
Jevin Tan
Business & Compliance
Fourteen years across healthcare delivery and regulation in Southeast Asia. Owns commercial, partnerships and the regulatory road map.
R
Raymond Harrison
AI Engineering
Built production AI for GSK, HTX and URA. Leads the pose detection, embeddings and inference stack that sits behind every SIMIS deployment.
J
Jenxi Seow
Operations & Hardware
Took products from prototype to volume manufacturing in Shenzhen. Owns the SIMIS box, the deployment workflow and on-site reliability.
Advisory: AI Singapore mentors and the wider Pan-SEA Developer Challenge network.
Get in touch
Ready to stop guessing what your cameras saw?
Three doors in. All of them open to the same small team.
Investment
We are mapping our next round. Happy to share the deck and walk you through the technical moat.
Open a conversation PartnershipCamera vendors, integrators, regulated-industry operators—if your customers ask the questions SIMIS answers, let us talk.
Explore a partnership PilotChildcare, eldercare, construction, infrastructure. If you operate sites where the cost of missing something is high, we would like to hear from you.
Discuss a pilot