About this role
Hands-on AI/ML engineering role to transform AssetHub’s real-world assets (codebases, workspaces, databases) into high-value AI-training products such as RL environments, agentic task suites, evals, verifiers, benchmarks, and datasets. Prototype quickly and harden solutions into repeatable pipelines, build reproducible sandboxes and tooling, and partner with buyers and lab stakeholders (candidates must be located in the New York City Metro area).
Skills for this role
RL environmentsAgentic task suitesEvaluations (evals)VerifiersBenchmarksFine-tuning datasetsTrajectory datasetsRepository ingestionTest harnessingCommit-miningDockerSandbox reproducibilityVerifier designReward designQA toolingSandboxed application stateMCPPlaywrightLLM evaluationAgent harnessesHUDPythonCITest infrastructureData privacyPII handlingSecurityLicensingPrototype developmentPipeline engineering''Simulation''Frontend development''Post-training data''