Sherpa Labs GTM brain

Sherpa Labs supplies frontier AI labs with expert-built reinforcement learning environments, benchmarks and training data for professional tasks.

category

Frontier AI training data and evaluations

market profile

  • text: Frontier model teams need challenging task environments and evaluations as generic data yields less improvement, according to Sherpa's thesis.
  • text: Finance and coding are identified by a vendor directory as target domains for Sherpa's environments.

personas

  • title: Head of post-training or reinforcement learning
    pains:
    • Generic datasets do not sufficiently challenge frontier models
    • Building realistic environments internally requires specialized expertise
  • title: AI evaluation research lead
    pains:
    • Benchmarks must represent professional workflows
    • Expert rubrics and hard tasks take time to design
  • title: Data partnerships lead at a model lab
    pains:
    • Sourcing qualified domain experts is operationally intensive
    • Data quality must withstand research scrutiny

icp

  • text: Frontier AI model developers investing in post-training and evaluation for professional knowledge work.
  • text: Teams pursuing coding or finance tasks are plausible early fits; specific account size and budget are undisclosed.

message pillars

  • text: Treat data creation as research.
  • text: Supply environments and benchmarks suited to professional work.
  • text: Engage leading practitioners in difficult domains.

objections

  • objection: How do we know tasks are sufficiently realistic?
    response: Review sample tasks, practitioner credentials and grading rubrics against the lab's held-out objectives.
  • objection: Can our research team build this in-house?
    response: Compare the time to source practitioners and construct reliable environments with an evaluation of Sherpa's sample.

competitors

  • name: Sepal AI
    why: Supplies expert-created training data, benchmarks and RL environments to model teams.
  • name: Surge AI
    why: Provides human data for model training and evaluation.

influencers

  • name: Justin Jiang
    url: https://xitter.gq/jostinjiang
    why: Founder posting about RL environments and training data.

keywords

  • reinforcement learning environments
  • frontier model training data
  • expert AI benchmarks
  • finance AI evaluation
  • coding agent evaluation
  • human preference data
  • RL post training data

voice

  • Research-led and technical
  • Confident and concise

signals

  • text: A model lab is hiring for RL post-training or domain-specific evaluations.
  • text: A team is expanding from general benchmarks into finance or coding workflows.

hiring

  • Strategic Project Lead, Finance
  • Strategic Project Lead, Coding
  • Research Engineer, RL Environments
  • Data Partnerships Lead
  • Growth Lead

pricing

model: unknown

motion

sales_led

channels

  • X
  • research benchmarks

confidence

medium

Founders

  • name: Justin Jiang
    title: Founder/CEO
    role: CEO
    bio: Building frontier RL data pipelines @ Sherpa Labs. Ex-research scientist from Surge, ex-quant researcher from Millennium, Vatic, Optiver, working on medium frequency statistical arbitrage and high frequency trading algorithms. Data scientist at FedEx and Tradeweb. Graduated from Harvey Mudd with a double BS in Math and CS.

Back to YC companies