Perit.AI GTM brain

Perit collects, annotates and licenses expert-produced speech and multimodal training data and RL tasks for AI labs and robotics teams.

category

AI training data and evaluation services

market profile

  • text: Frontier model teams need specialized human-produced training and evaluation data beyond generic clean speech corpora.
  • text: Speech, RL and embodied AI programs require diverse collection, annotation and documented quality at scale.

personas

  • title: Speech model training lead
    pains:
    • Clean read speech misses realistic interruptions and domain-specific call errors.
    • Low-resource locales and word-level alignment are difficult to source consistently.
  • title: Robotics data lead
    pains:
    • Needs demonstrations in varied real homes and workplaces, captured to specific rig and format requirements.
    • Collection must include defensible consent, annotation and quality checks.
  • title: AI data procurement or operations lead
    pains:
    • Needs predictable throughput and visibility into quality, licensing and contributor pay.
    • Data residency and exclusivity requirements complicate vendor evaluation.

icp

  • text: Frontier AI labs training speech, multimodal or reinforcement learning models with specialized human data needs.
  • text: Robotics and physical AI teams gathering egocentric video, depth and human demonstrations in real environments.
  • text: Teams running multilingual voice programs where locale, transcript alignment and consent materially affect data utility.

message pillars

  • text: Use people who do the work to record, transcribe and grade domain-specific data.
  • text: Scale speech production across nine locales with quality reporting on each batch.
  • text: Specify physical-world data collection, including demonstrations, depth and customer-supplied rigs.
  • text: Make consent, license, contributor pay and measured quality visible in the program.

objections

  • objection: Can you meet our quality bar for a specialized locale?
    response: Scope a pilot and inspect the written guide, calibration and batch QA report before scaling.
  • objection: Can you handle sensitive or proprietary data?
    response: The pricing page offers customer cloud and agreed regional handling; validate contractual and security terms for the program.
  • objection: What will the program cost?
    response: Request a scoped quote; locale, turnaround, annotation, exclusivity and capture setup affect price.

competitors

  • name: Appen
    website: https://www.appen.com/
    why: Provides multilingual speech, multimodal, RL and physical AI data services.
  • name: Scale AI
    why: Managed training data and evaluation for AI labs.
  • name: Surge AI
    why: Expert annotation and reinforcement learning data provider.
  • name: iMerit
    why: Managed specialist annotation for multimodal and regulated use cases.

keywords

  • speech training data collection
  • multilingual speech transcription datasets
  • word alignment annotation
  • robotics egocentric video data
  • robot demonstration datasets
  • RL training environments
  • expert AI data annotation
  • licensed AI training data

voice

  • Operational and specific
  • Evidence-led about quality and provenance
  • Plainspoken about scope and tradeoffs

signals

  • text: A lab expanding a speech model to new locales or identifying failures on real calls is a strong fit.
  • text: A robotics team collecting new manipulation demonstrations or specifying a capture rig is a plausible trigger.
  • text: A data buyer seeking documented consent and quality for a new training run has a clear reason to evaluate a pilot.

stack

  • Foundry
  • Hugging Face

hiring

  • AI Data Annotation (Fixed Term)
  • Freelance AI Data Annotation (Alignment)
  • Audio Conversation Rater (Fixed Contract)

pricing

model: contact_sales

motion

sales_led

channels

  • SEO
  • research benchmarks
  • open datasets
  • LinkedIn

confidence

high

Founders

  • name: Kshitij Dixit
    title: Founder/CEO
    role: CEO
    bio: Co-founder & CEO at Perit.AI. Previously co-founder/CEO at ZeoAuto (YC W20), scaled to millions of drivers across 140+ countries, $2M ARR profitable, 4.8 stars from 30k+ reviews. Led consumer payments at Jio Financial Services with 10M+ users. Former Senior Product Manager at OYO. IIT Kharagpur.
  • name: Poorva Dixit
    title: Founder/CTO
    role: CTO
    bio: Co-founder & CTO at Perit.AI. Previously Software Engineer at Google, Atlassian and Uber. Co-founded SalaryBook. Dual degree in Computer Science from IIT Kharagpur.

Back to YC companies