What it sells
Analyze any footage, without training a model
Business model breakdown
Buyer: who owns the budget?
Defence agencies, public safety, robotics and enterprise operations/safety teams with large video archives.
Trigger: what makes them buy now?
UnverifiedPetabytes of unstructured footage that analysts cannot review manually, and off-the-shelf vision-language models failing on unseen data.
Workflow: what inputs become what output?
Describe what to find in plain language (cracks, policy violations, anomalies), OnDeck searches footage and telemetry, explains cause, produces timestamped clips and human or machine-readable incident reports, and triggers alerts.
Economics: what it saves and how it charges
Enterprise vision model licensing with cloud or on-prem/air-gapped deployment and per-customer customization; SOC 2 certified; claims Perception-0 scores 81.2% on LVBench vs 78.5% for Gemini; Singapore Ministry of Defence cited as customer.
Distribution: how it reaches customers
Sales-led ('Talk to our team', book a call) by industry vertical: enterprise, defence, environmental.
Funding history and coverage
| Round date | Stage | Amount | Investors | Source |
|---|---|---|---|---|
| — | Y Combinator Summer 2025 | — | Y Combinator | ycombinator.com |
Profiles summarize public reporting and company websites. Business-model notes are MaxBC's reading of that material, not confirmed facts; unknowns are left blank rather than guessed. Updated 2026-10-12.