Agent Harness Engineer

Axiom; San Francisco (CA); Hybrid; USD 180,000 - 260,000; Full time

Job summary

Axiom is building a compounding ecosystem to replace animal testing and reshape how clinical trials are run. We focus on drug‑induced liver injury and leverage a data and agentic harness to power ML research.

You will own the harness, pipelines, and infrastructure that turn frontier models into reliable agents, and you’ll work with domain experts to codify evaluation rubrics. This is a hands‑on role for engineers who ship clean, scalable systems and stay ahead of state‑of‑the‑art methods in AI

Qualifications

• Owns and ships end‑to‑end systems for agentic workflows.
• Architects data pipelines, storage, and runtime context for agents.
• Build sandboxed, deterministic environments for evals and RL.
• Design and run offline eval suites and LLM‑as‑judge pipelines.
• Collaborates with domain experts to encode rubrics and review workflows.
• Keeps agent runs observable with traceable model/tool/state transitions.
• Develops memory, retrieval, and recovery to maintain coherence over long horizons.
• Owns retry loops, budget caps, stop conditions, and guardrails.
• Supports ML research with environments and rollout infra for RL tasks.

Responsibilities

• Own the harness, tooling, and infrastructure for agent systems.
• Build the data backbone: pipelines, storage, contexts, and agent trajectories.
• Create sandboxed environments with instant spin‑up/tear‑down and deterministic evals.
• Design/run eval systems: offline suites, LLM‑as‑judge, regression gates.
• Encode expert taste into rubrics, golden sets, and review workflows.
• Build tools the agent needs and stay updated with new methods and patterns.

Skills

Python

Modal

DuckDB

FastAPI

Docker

Containerization

Terraform

LLM APIs

Tools

SvelteKit

Svelte 5

React

Job description

About Axiom:

Axiom is building a compounding ecosystem to replace animal testing and, over time, reshape how clinical trials are run. It starts with deeply understanding the needs of drug hunters inside large pharma. Those needs shape the world‑class datasets we build from scratch. We then use that data to advance our own ML research, while also collaborating with leading AI labs to improve frontier models’ ability to reason over Axiom’s data inside Axiom’s agent harness. This creates a compounding loop: deeper customer understanding shapes the data we generate; better data improves frontier models, Axiom’s fine‑tuned models, and our agentic infrastructure; stronger models and tooling expand the capabilities we can offer; and those capabilities are forward deployed into pharma’s drug discovery workflows, where scientists use them to solve the highest value drug discovery problems. In turn, this helps us identify the next problems to tackle. Today, we are focused on solving drug‑induced liver injury through an integrated data and agentic system already being used by 7 of the top 20 pharma companies and several of the world’s most innovative biotechs. Over time, Axiom will build the world’s largest human datasets across all the major organ systems, paired with an agentic harness that uses this data to predict human drug outcomes dramatically better than animals.

What you will do:

• Own the harness: the scaffolding, tooling, and infrastructure that turn frontier models into agents that do long‑horizon scientific analysis
• Build the data backbone that gets everything to the right place: pipelines, storage, and systems for runtime context, agent trajectories, eval results, and training data
• Build sandboxed execution environments with instant spin‑up/tear‑down, reproducible and deterministic enough to trust for evals and RL
• Design and run the eval systems: offline suites, test cases on production traces, LLM‑as‑judge pipelines, regression gates
• Work closely with domain experts and encode their taste into rubrics, golden sets, and review workflows, turning “I know it when I see it” into something measurable
• Build the tools the agent needs to do better work, and stay tuned in to the state of the art for new methods, protocols, and patterns worth adopting
• Make every agent run observable and replayable: trace every model call, tool call, and state transition, and build the debugging tooling to make sense of it
• Engineer the context: memory, compaction, retrieval, and recovery so long‑horizon agent runs stay coherent across hours and crashes
• Own the loop itself: retries, budget caps, stop conditions, output verification, permissions, and guardrails
• Support ML research with environments, reward instrumentation, and rollout infra for RL on agentic tasks

Expertise:

• Python, Modal, DuckDB, FastAPI, Docker, Containerization, Terraform
• Engineers who've built with LLM APIs and shipped agentic systems: tool use, loops, and the debugging scars to prove it
• Built bespoke evaluation, monitoring, and RL env observability tooling (SvelteKit, Svelte 5, React)

What we look for:

• Can tackle deep technical challenges and own/ship simple, clean, maintainable code
• High ownership: owns outcomes end to end, not tickets, and doesn't wait for a spec to start moving
• Allergic to complexity: reaches for the simplest system that works and keeps it that way as it scales
• Strong software engineer first, with infrastructure, platform, data, or devtools depth and production systems they're proud of
• Instinctively asks "how would we know if this is working?" and builds the measurement alongside the feature
• Reads an agent failure trace the way other engineers read a stack trace
• Cares about reliability because they know environments that break silently poison evals and training data
• Has a knack for surfacing the important questions about what the agent actually needs to do better work
• Sharp and confident
  • keeps up with a really technical & sophisticated crowd
  • comfortable getting in over their head and figuring it out as they go
  • thrives in a discipline with no playbook, because it's being invented right now
• Curious about how things work: an engineering/tinkering mindset, good at scavenging the state of the art
• Passion for learning what "good" looks like from deep domain experts and turning it into systems