OnDeck AI GTM brain

OnDeck AI sells enterprise vision-language infrastructure that searches and interprets video without building a task-specific model.

category

Enterprise video intelligence

market profile

  • text: Enterprise teams must analyze large video archives and events across differing cameras and environments.
  • text: OnDeck positions Perception-0 against general-purpose video-capable models on long-video benchmarks.

personas

  • title: Defense intelligence and situational awareness lead
    pains:
    • Analysts cannot manually review every camera stream and autonomous asset recording.
    • Sensitive environments may require air-gapped deployment.
  • title: Robotics research lead
    pains:
    • Finding specific behaviors in footage is labor intensive.
    • Custom vision training requires scarce labeled examples and may not generalize.
  • title: Security operations director
    pains:
    • Safety events are buried in large unstructured video collections.
    • Different sites need different detections and workflows.

icp

  • text: Defense agencies and public safety organizations with large video estates and strict deployment requirements.
  • text: Robotics companies and research institutions seeking fine-grained behavior analysis.
  • text: Enterprise operators with multiple camera environments and repeatable video review workflows.

message pillars

  • text: Search large volumes of video for relevant events.
  • text: Interpret complex events rather than only locating objects.
  • text: Fit the model to each customer's workflow.

objections

  • objection: Can sensitive footage stay in a controlled environment?
    response: The company lists on-premises and air-gapped deployment options; validate the deployment for the specific environment.
  • objection: Will the model work on our unusual footage?
    response: Pilot with representative footage and measure event-level recall and precision; OnDeck says it adapts to customer workflows.

competitors

  • name: Google Gemini
    why: General-purpose vision-language models can analyze video and appear in OnDeck's benchmark.
  • name: Microsoft Azure
    why: Azure Deep Video Discovery is listed as a video understanding benchmark alternative.
  • name: OpenAI
    why: GPT-4o is another multimodal model listed in the company's comparison.

keywords

  • video intelligence platform
  • zero shot video analytics
  • vision language model video search
  • defense video analytics
  • robotics video understanding
  • air gapped computer vision

voice

  • Technical and performance-led
  • Ambitious
  • Enterprise-oriented

signals

  • text: A defense or security team is expanding video surveillance and cannot manually review footage at scale.
  • text: A robotics team is collecting video of rare behaviors and needs flexible retrieval.

hiring

  • Software Engineer, Vancouver
  • Machine Learning Researcher, Vancouver

pricing

model: unknown

motion

sales_led

channels

  • Website
  • LinkedIn
  • X
  • Research publications
  • Direct sales

confidence

medium

Founders

  • name: Alexander Dungate
    title: Founder/CEO
    role: CEO
    bio: Solving vision @ OnDeck. BSc in CS + biology, National Geographic Explorer.
  • name: Sepand Dyanatkar
    title: Founder/CTO
    role: CTO
    bio: Solving vision. Masters @ Cambridge in multi-agent RL. Research in swarm robotics for space, and built software at Amazon and the European Space Agency.

Back to YC companies