Startups · AI

Platform Research Engineer

Applied Compute · San Francisco · On-site

← All jobs
About Applied Compute

Specific Intelligence for Your Business. Backed by Kleiner Perkins and Lux.

About the role

As a platform research engineer, you'll do research around ways to make Applied Compute's platform intelligent. This includes how we set up continual learning against production traces (OPSD, SDPO, RMSD), scaling synthetic environments, data & evaluations, and integrating the RL stack that powers agent improvement over time.

What they're looking for

  • Strong software engineering fundamentals combined with deep ML/AI knowledge
  • Production research experience: both deliveries train models into production & top-tier conference publications, blogs, or reports
  • Strong experimental design skills: you are diligent, don't cut corners, and actually run experiments
  • Highly organized: you manage complexity across multiple workstreams
  • Ability to read and translate research papers and prototypes into shippable engineering
More about this role

As a platform research engineer, you'll do research around ways to make Applied Compute's platform intelligent. This includes how we set up continual learning against production traces (OPSD, SDPO, RMSD), scaling synthetic environments, data & evaluations, and integrating the RL stack that powers agent improvement over time. Your research will serve as the connective tissue between our AI product engineers (who build the interfaces and tools customers use) and our applied research engineers (who work directly with customers to ship agents into production). Your job is to take learnings from across customer deployments, identify the greatest common denominators, and pursue research to accelerate all of our deliveries.

What you'll do

Build and improve our hinting primitives for online self-distillation: the algorithms and systems that allow agents to learn continuously from production traces and company data

Develop and maintain trace search, trace mining, and data labeling systems that feed the continual learning loop

Research and implement approaches to multi-agent system design, including structuring agent coordination and managing trade-offs

Build best in class coding agents for...

Read the full posting on Applied Compute's site ↗

Platform

Build your edge while you search

Free tools for founders and investors, plus VC Unfiltered, our take on startups, venture and the people who build them.