/jobs

Senior Harness Engineer

Last updated:

The model is only one part of an agent. The harness around it — its tools, context, memory, execution loop, safety boundaries, and feedback mechanisms — often determines whether it succeeds outside a demo.

We're looking for a senior Python engineer to own and grow Pydantic AI Harness, our open source library of reusable capabilities for building production-grade agents with Pydantic AI. These capabilities bundle tools, lifecycle hooks, instructions, and model settings so developers can add sophisticated agent behavior without rebuilding their agent or changing the framework.

This is a hands-on product and open source engineering role. You'll build the components that let agents work effectively over long-running, complex tasks; use Logfire traces and rigorous evaluations to understand where they fail; and turn those findings into measurable improvements. You'll also help shape Pydantic AI's core extension mechanisms when the work calls for it.

Harness engineering is still a young discipline. We want someone who has already learned the hard lessons of shipping agents in production and is excited to define what good looks like in public.

  • Own the technical direction and day-to-day development of Pydantic AI Harness, working primarily in Python.
  • Design, build, and maintain reusable agent capabilities such as context management and compaction, memory, planning, skills, sub-agents, steering, sandboxed tool execution, guardrails, monitoring, and recovery from failed or stuck runs.
  • Build with the model as a user of the system: make state, provenance, errors, limits, and changes to its environment legible so it can act on an accurate view of the world.
  • Dogfood the harness in real agents and agentic development workflows, then turn what breaks into better APIs, capabilities, documentation, and defaults.
  • Instrument agents with Pydantic Logfire, analyze production traces, create evaluations and benchmarks, and verify that changes improve task success, reliability, latency, and cost rather than merely looking plausible.
  • Establish reproducible benchmarks against leading agent harnesses and keep pace with a fast-moving field through experiments, competitor research, and close attention to new models and techniques.
  • Work across Pydantic AI and the harness: improve the capabilities and hooks extension APIs in core when a clean harness implementation needs framework support.
  • Maintain a high-quality open source project: respond to users, review issues and pull requests, help contributors turn useful ideas into composable capabilities, and communicate API changes clearly.
  • Ship quickly while preserving trust through strong tests, useful deprecation paths, clear documentation, and evidence from real agent runs.

We expect a candidate for this position to have:

  • Significant experience building, operating, and improving LLM agents or agent harnesses in production, particularly agents that work autonomously over long-running, complex tasks. You know the difference between an impressive prototype and a system that remains useful under real workloads.
  • Strong Python engineering skills and hands-on experience with Pydantic AI or another agent framework. You understand agent loops, tools, context management, structured outputs, model-provider differences, and the abstractions frameworks use to bring them together.
  • Experience instrumenting production agents with an LLM observability platform, diagnosing their behavior from traces, and using that evidence to improve them. You can turn a vague instruction like “make the agent better” into hypotheses, experiments, and measurable results.
  • Hands-on experience with the systems around a production agent: sandboxed code or tool execution, permissions and safety boundaries, long-running state, and multi-agent orchestration.
  • Experience connecting agents to external tools and services through MCP or comparable tool and context protocols.
  • End-to-end ownership. You can identify the important problem, research the landscape, design the right abstraction, implement it, document it, and validate it with users without waiting for a detailed specification.
  • Good judgment about the interaction between model behavior and software design. You understand that prompts, tool schemas, errors, context management, and execution semantics all shape what an agent can reliably do.
  • An interest in building open source software and working directly with a technical community in public.
  • Comfort using coding agents as a normal part of software development while retaining responsibility for architecture, correctness, and quality.
  • At least 5 years of software engineering experience.

Nice to haves but not required:

  • Experience designing agent evaluation suites or working with benchmarks such as Terminal-Bench, SWE-bench, or similar task-based evaluations.
  • Contributions to open source agent tooling, Python developer tooling, SDKs, or frameworks.
  • Experience maintaining an open source project or helping its contributor community grow.
  • Familiarity with OpenTelemetry and Pydantic Logfire specifically.
  • Experience with ACP or other protocols for connecting agents to user interfaces.
  • Live and work in a timezone between PT (UTC-8) and CET (UTC+1)
  • Able to travel to the EU, UK, and US up to 4 times a year to join our off-sites

Pydantic Validation is the data validation library that powers modern Python development - 500 million downloads per month, used by virtually every tech company you've heard of. Why? Because we obsess over developer experience and write code we'd actually want to use ourselves.

We're applying that same engineering mindset to Pydantic Logfire, our observability platform with first class support for AI engineering, built for today's development reality: AI workloads, multi-language environments, and cloud infrastructure that's designed to be straightforward to set up and maintain.

We build with technologies developers actually want to work with:

  • OpenTelemetry for standardized instrumentation
  • SQL for intuitive querying (no proprietary query language to learn)
  • Rust, Python, and TypeScript for performance and productivity
  • Postgres, DataFusion, and object storage for scalable backends

Unlike other companies that pay lip service to open source, we commit over 20% of our engineering team to maintaining and expanding our open source ecosystem. This includes the core Pydantic Validation library and Pydantic AI - our rapidly growing framework that's becoming the standard for AI application development. We're signatories of the open source pledge and build on open standards because we believe in interoperability, not lock-in. Use our OpenTelemetry-based SDK with any compatible backend - we're confident you'll choose us on merit.

We're backed by Sequoia Capital and run a fully remote team across multiple time zones (with regular in-person offsites - next one is June 2026 in London).

Join our team of exceptional engineers who value substance over hype, practical approaches over perfectionism, and meaningful progress over busyness. We've built a culture that balances technical ambition with sustainable practices—minimal meetings and respect for your expertise and time. We're creating tools that genuinely improve developers' lives, and we're looking for thoughtful contributors who share our commitment to quality and our passion for elegant solutions.

  • 💰 Compensation: Competitive salary and stock options
  • 🌍 Truly Remote: Work from anywhere within our timezone range - no office requirements
  • 🌐 Global & Diverse: Join a multi-cultural team of 8+ nationalities
  • 💪 Impact: Direct influence on tools used by millions of developers worldwide
  • 🎯 Focus on Growth: Regular opportunities for learning and professional development
  • 🤝 Team Gatherings: Connect with the team at our regular international off-sites
  • 🏥 Healthcare: Comprehensive health coverage for you and your dependents
  • 🎮 Flexible Hours: Work when you're most productive
  • 💻 Equipment: Budget for your home office setup
  • ⚖️ Work-Life Balance: flexible working hours and 33 days PTO no matter where you live (including public holidays, which you can choose to take or not)

To apply, email careers@pydantic.dev with the job title in the subject line. We'd also appreciate a few lines explaining why you think you'd be a good fit for the role and what you've done in the past that evidences that.

No recruiters or agencies please. Unsolicited recruiters will be marked as spam.

To make your application stand out, please share an agent or harness you've built and run in production, an open source contribution, or a technical write-up that shows how you measured and improved agent behavior.

Explore Logfire.

Sign up

Explore our open source packages

Product shot