---
title: "Research Engineer, Agentic Coding Environments"
slug: "research-engineer-agentic-coding-environments"
id: "0e92c376-facb-4394-9ca1-9299591a2af4"
department: "Machine Learning"
department_code: "ML"
location: "San Francisco, CA"
workplace_type: "Hybrid"
employment_type: "Full-time"
compensation: "$260k – $360k (annual package)"
is_remote: false
is_featured: false
headcount: 2
canonical_url: "https://jobs.talent247.co/jobs/research-engineer-agentic-coding-environments"
markdown_url: "https://jobs.talent247.co/jobs/research-engineer-agentic-coding-environments.md"
apply_endpoint: "https://jobs.talent247.co/api/public/jobs/research-engineer-agentic-coding-environments/apply"
schema_endpoint: "https://jobs.talent247.co/api/public/jobs/research-engineer-agentic-coding-environments/form-schema"
webmcp_tool: "apply_for_role"
updated_at: "2026-11-15T17:38:06.207Z"
content_signal: "search=yes, ai-input=yes, ai-train=no"
---

# Research Engineer, Agentic Coding Environments

## Position Telemetry

- **Reference:** `RESEARCH-ENGINEER-`
- **Practice / Sector:** Machine Learning (`ML`)
- **Location:** San Francisco, CA
- **Workplace Type:** Hybrid
- **Employment Type:** Full-time
- **Compensation:** $260k – $360k (annual package)
- **Headcount:** 2 openings
- **Canonical HTML URL:** https://jobs.talent247.co/jobs/research-engineer-agentic-coding-environments

## Role Dossier & Specification

THE CLIENT & MANDATE
Our client is a premier San Francisco frontier AI laboratory whose flagship models set the industry benchmark for autonomous software engineering, terminal orchestration, and long-horizon computer use. Behind those coding breakthroughs sits an elite Agentic Coding Environments research engineering group building the realistic, multi-hour software sandboxes and reinforcement learning curricula where frontier models learn to navigate, architect, debug, and verify production-grade codebases.

They have engaged Talent247 to hire a Research Engineer for Agentic Coding Environments in San Francisco, CA. Bridging RL research and systems engineering, the successful candidate will build massively scalable repository gyms, compiler/execution verifiers, and multi-turn post-training loops pushing autonomous coding capabilities from single-PR patches to complex multi-service engineering projects.

KEY RESPONSIBILITIES
• Architect hermetic, sub-second containerized and microVM coding environments supporting polyglot build systems (Rust, C++, Go, TypeScript, Python, Java, Kubernetes, databases) across tens of thousands of parallel RL rollouts.
• Design rich, cheat-proof verification harnesses combining hidden integration tests, property-based fuzzing, performance/memory regression benchmarks, and calibrated code-review reward models.
• Run large-scale multi-turn reinforcement learning experiments teaching models long-horizon repository exploration, hypothesis-driven debugging, git history archaeology, and clean architectural refactoring.
• Build automated task-generation pipelines that synthesize realistic feature migrations, subtle concurrency bugs, and security patches across complex multi-repo topologies without benchmark contamination.
• Analyze long-context (100k+ token) agent trajectories to improve context-compaction habits, shell/editor tool ergonomics, and self-correction when builds fail.
• Partner closely with developer-product and safety teams to ensure autonomous coding agents respect sandbox boundaries and avoid destructive operational actions.

TECHNICAL ENVIRONMENT & OFFERING
Stack: Python, Rust, TypeScript, PyTorch/JAX, Firecracker/Docker/Kubernetes rollout clusters, LSP/tree-sitter/Bazel/Nix build toolchains, frontier GPU/ASIC training clusters. Workplace: Hybrid in San Francisco, CA. Compensation: $260,000 – $360,000 base salary plus substantial equity grants, full healthcare, and generous benefits.

## Core Requirements

ESSENTIAL QUALIFICATIONS
• 4+ years of software systems and ML/RL research engineering experience with exceptional polyglot coding craftsmanship.
• Hands-on experience building coding-agent harnesses, SWE-bench-style containerized evaluation/RL gyms, or large-scale LLM post-training pipelines.
• Deep systems fluency in Linux containerization, microVM snapshotting, reproducible polyglot builds (Docker, Nix, Bazel), and distributed task scheduling.
• Strong intuition for how language models interact with shell terminals, file editors, debuggers, and test runners—and how they attempt to game naive graders.
• Ability to own both the systems reliability of 100k-container rollout fleets and the scientific analysis of RL training dynamics.

PREFERRED / BONUS EXPERIENCE
• Prior experience training frontier or open-weight coding models with online RL (GRPO/PPO) over multi-turn interactive environments.
• Background in compilers, static analysis, fuzzing, OS internals, or developer tools engineering.
• Track record of high-impact open-source contributions or strong competitive programming achievements.

## Application Schema & Direct Apply (`apply_for_role`)

Candidates and authorized AI agents on behalf of a candidate can submit an application via the declarative WebMCP form tool `apply_for_role` on https://jobs.talent247.co/jobs/research-engineer-agentic-coding-environments or via HTTP `POST https://jobs.talent247.co/api/public/jobs/research-engineer-agentic-coding-environments/apply` (`application/json` or `multipart/form-data`). Non-browser HTTP agents must include header `X-Talent247-Intake-Token` obtained from `GET https://jobs.talent247.co/api/public/intake-token` (or automatically returned in `GET https://jobs.talent247.co/api/public/jobs/research-engineer-agentic-coding-environments/form-schema`).

### Form Fields

- `full_name` (text, required): Full name
- `email` (email, required): Email
- `phone` (phone, optional): Phone
- `location` (text, optional): Location
- `linkedin_url` (url, optional): LinkedIn URL
- `cover_letter` (textarea, optional): Cover letter
- `consent` (checkbox, required): I consent to my data being processed for this application
- `cv_file` / `cv_base64` + `cv_filename`: Upload a CV in .pdf or .docx format (<= 10 MB) or provide an `https://` LinkedIn profile URL in `linkedin_url`.

### Alternative: Talent Pool Intake (`join_talent_pool`)

If this mandate is not an exact match, register once in the Talent247 Talent Pool at https://jobs.talent247.co/talent-pool (`https://jobs.talent247.co/talent-pool.md`) via WebMCP tool `join_talent_pool` or `POST https://jobs.talent247.co/api/public/talent-pool`.
