# OnDeck AI

> OnDeck AI is a visual intelligence research company building the world's most
> powerful vision model. Our models solve complex video analysis tasks with no
> training and no data labelling - find visual behaviours, structural damage,
> and anomalies in footage in seconds.

OnDeck is deployed with defence agencies, universities, robotics companies, and
public safety and environmental organizations. Models can run air-gapped,
on-prem, and private-cloud for complete data sovereignty. SOC 2 Type II
compliant. Founded 2022. Based in Vancouver, Canada. Backed by Y Combinator.

## Flagship model: Perception-0

Perception-0 is not a model in the traditional sense - it is a self-improving
harness that extracts state-of-the-art performance from frontier models and
autonomously improves on each task. As the underlying frontier models improve,
Perception-0 keeps getting better.

Benchmark - LVBench (long-video understanding), overall score:

- Perception-0: 81.2
- Gemini 3.1 Pro: 78.5
- Deep Video Discovery: 74.2
- Seed1.5: 64.6
- AdaReTaKe: 53.3
- GPT-4o: 48.9

## Proof points

- Singapore Ministry of Defence (MINDEF): threat detection and situational
  awareness across autonomous assets.
- Parametric (robotics): fine-grained behaviour understanding to accelerate
  training of safe robotics models.
- SOC 2 Type II compliant; air-gapped / on-prem / private-cloud deployment.
- Backed by Y Combinator.

## Pages

- [Home](https://ondeckai.com/): The world's most powerful vision model - capabilities, benchmarks, and research.
- [Thesis](https://ondeckai.com/thesis): Why adaptive visual intelligence is the path to human-level perception.
- [Enterprise](https://ondeckai.com/industries/enterprise): Extract everything from your footage without training a model.
- [Defence](https://ondeckai.com/industries/defence): The most powerful visual intelligence model for ISR, deployed with full data sovereignty.
- [Environmental](https://ondeckai.com/industries/environmental): 100x faster visual monitoring for environmental organizations.
- [Careers](https://ondeckai.com/careers): Join OnDeck to build the final generation of visual intelligence.

## Research

- Peer-reviewed research published at NeurIPS. Paper: https://arxiv.org/abs/2412.02262 -
  improving vision-language-model reliability on unseen data by grounding
  predictions in real data rather than model guesses.

## Contact

- Talk to our team: https://cal.com/team/ondeck-ai/intro
- LinkedIn: https://www.linkedin.com/company/ondeckai/
- Twitter/X: https://x.com/OnDeckAI