shotwell.ai GTM brain

Shotwell AI labels and quality-checks robotics demonstration and deployment data for teams training dexterous robots.

category

Robotics data annotation and QA

market profile

  • text: Robotics labs, deployment companies and data vendors need usable labels and quality checks on recorded robot tasks.
  • text: Physical AI data suppliers range from specialist annotation vendors to platforms offering collection and multimodal labeling.

personas

  • title: Head of robot learning
    pains:
    • Training examples lack consistent subtask boundaries and semantic labels
    • Slow inspection delays model iteration
  • title: Robotics data operations lead
    pains:
    • Reviewers must inspect large volumes of teleoperation footage
    • Task failures and deviations from standard procedures are difficult to locate
  • title: Robotics data vendor lead
    pains:
    • Customers need auditable task quality and portable label outputs
    • Multiple capture formats complicate delivery

icp

  • text: Robot learning labs collecting dexterous manipulation demonstrations through teleoperation or egocentric video are a natural fit.
  • text: Robotics deployment companies and data suppliers with enough recurring video to make manual review a bottleneck are promising accounts; no firm size is stated.

message pillars

  • text: Make every episode understandable and reviewable.
  • text: Deliver labels quickly in practical formats.
  • text: Find failure paths as well as successful subtasks.

objections

  • objection: Will labels match our task definition?
    response: Provide the SOP and trial a representative batch, checking boundaries and rubric verdicts against expert review.
  • objection: Can we integrate outputs into our pipeline?
    response: The site lists MP4, MCAP and LeRobot inputs and JSON or CSV outputs; confirm schemas and security requirements in a pilot.

competitors

  • name: Scale AI
    why: Provides physical AI data collection and multimodal annotation services.
  • name: Labelbox
    why: Offers robotics data and multimodal annotation workflows.
  • name: Encord
    why: Offers physical AI data management and annotation tooling.

keywords

  • robotics data annotation
  • teleoperation video labeling
  • robot learning data quality
  • egocentric data annotation
  • robot task segmentation
  • robotics failure detection

voice

  • Technical and concrete
  • Evidence-led
  • Pragmatic

signals

  • text: A team is scaling teleoperation or egocentric demonstration collection and has a growing annotation backlog.
  • text: A robot lab needs to find task failures or evaluate operator adherence to a documented SOP.

stack

  • MCAP
  • LeRobot
  • JSON
  • CSV

pricing

model: unknown

motion

hybrid

channels

  • SEO
  • product demo
  • robotics case studies

confidence

medium

Founders

  • name: Farhan Khan
    title: Founder/CEO
    role: CEO
    bio: Previously worked on humanoid robots at Sunday Robotics and self-driving cars at Tesla
  • name: Ali Abdalla
    title: Founder
    role: Founder
    bio: Previously: co-founded Gradio which was acquired by Hugging Face and used by >1M developers a day, Tesla, MIT.
  • name: Ali Abid
    title: Founder/CTO
    role: CTO
    bio: Previously found Gradio (acq. by HuggingFace) Now interested in robotics.
  • name: Nour Eldifrawy
    title: Founder
    role: Founder
    bio: Currently working on unlocking robotics teams. Formerly @ Merge.dev, Roblox, Tarjimly (yc w18)

Back to YC companies