Staff Machine Learning Engineer

People In AI; San Francisco (CA); Hybrid; USD 265,000 - 280,000; Full time

Benefits offered by this job

Hybrid work model

Equity

Job summary

People In AI in San Francisco is seeking a Staff-level Machine Learning Engineer to join the rapidly growing ML team. You will work on reinforcement learning, model post-training, and evaluation, bridging research and production systems.

The role emphasizes hands-on engineering, technical leadership, and scalable infrastructure. You will design architectures, run large-scale experiments, and guide a small team through ambiguous problems while shipping robust ML workflows.

Qualifications

• Hands-on ML engineering experience beyond model deployment or API integration.
• Experience with model fine-tuning or post-training workflows.
• Experience with reinforcement learning, evaluation, and training environments.
• Strong ability to lead technically without formal authority.
• Comfort in ambiguous, rapidly changing environments.
• Excellent Python and production software fundamentals.

Responsibilities

• Design and build reinforcement learning environments for agentic tasks.
• Develop task definitions, tool interfaces, state management, reward structures, and evaluation logic.
• Build systems capable of running large numbers of agent trajectories and experiments in parallel.
• Develop verifiers, graders, rubrics, and other evaluation methods for open-ended model behavior.
• Build fine-tuning and post-training pipelines spanning supervised fine-tuning and reinforcement learning.
• Run ML experiments and diagnose model behavior, training performance, reward quality, and data quality.
• Build scalable evaluation systems to measure model and agent performance.
• Develop production-grade ML infrastructure, including orchestration, reliability, fault tolerance, and experiment management.
• Translate ambiguous technical problems into clear architectures and execution plans.
• Provide technical leadership to other engineers while remaining hands-on.
• Use modern AI development tools while maintaining engineering judgment around the systems they produce.

Skills

Hands-on ML Engineering

Python

Distributed Systems

Production-grade Software Engineering

Technical Leadership

Model Evaluation & Post-Training

Experimentation & R&D

Job description

$265,000 - $280,000 Base Salary + Equity

Location: San Francisco – hybrid, 3 days per week in office

The Company

A fast-growing AI technology company is expanding its Machine Learning team in San Francisco.

The business builds advanced systems used to train, evaluate, and improve modern AI models. Its work spans reinforcement learning, post-training, model evaluation, agentic systems, and the infrastructure required to support increasingly sophisticated AI capabilities.

The engineering environment is highly technical and moves quickly, with a strong emphasis on ownership, judgment, experimentation, and execution.

The Opportunity

This is a Staff-level Machine Learning Engineering role sitting between research engineering and production ML infrastructure.

You will work on technical problems involving reinforcement learning, model post-training, agent training, evaluation, and the systems required to run these workflows reliably at scale.

This is not an application‑layer role focused primarily on integrating third‑party LLM APIs. The work sits closer to the underlying models, training signals, environments, evaluation methodologies, and infrastructure that determine whether AI systems actually improve.

The Role

You will combine hands‑on Machine Learning Engineering with Staff-level technical leadership.

The role is highly dynamic. Depending on the project, you could be running post‑training experiments, designing evaluation systems, building ML infrastructure, defining technical architecture, or leading a small group of engineers through an ambiguous technical problem.

The team already has strong implementers. This hire is intended to add another level of technical judgment and direction, helping determine what should be built, how it should be designed, and how the team should execute.

What You’ll Do

• Design and build reinforcement learning environments for agentic tasks.
• Develop task definitions, tool interfaces, state management, reward structures, and evaluation logic.
• Build systems capable of running large numbers of agent trajectories and experiments in parallel.
• Develop verifiers, graders, rubrics, and other approaches for evaluating open-ended model behavior.
• Build fine‑tuning and post‑training pipelines spanning supervised fine‑tuning and reinforcement learning.
• Run ML experiments and diagnose model behavior, training performance, reward quality, and data quality.
• Build scalable evaluation systems used to measure model and agent performance.
• Develop production‑grade ML infrastructure, including orchestration, reliability, fault tolerance, and experiment management.
• Translate ambiguous technical problems into clear architectures and execution plans.
• Provide technical leadership to other engineers while remaining deeply hands‑on.
• Use modern AI development and coding tools while maintaining strong engineering judgment around the systems they produce.

What You’ll Bring

• Strong hands‑on Machine Learning Engineering experience with meaningful depth beyond model deployment or API integration.
• Practical experience with model fine‑tuning or post‑training.
• Exposure to supervised fine‑tuning and reinforcement learning or preference‑optimization techniques such as GRPO, PPO, DPO, or similar.
• Experience with agentic systems, model evaluation, training environments, reward design, or closely related areas.
• Strong understanding of how to determine whether a model is improving and how to diagnose unsuccessful training runs.
• Strong Python skills and excellent production software engineering fundamentals.
• Experience with system design, distributed systems, ML infrastructure, platform engineering, or data systems.
• Strong architectural judgment and the ability to make and defend difficult technical decisions.
• Experience providing technical direction to other engineers without relying on formal management authority.
• High agency and a track record of identifying important problems independently and driving them through to completion.
• Comfort working in ambiguous, rapidly changing environments.
• Strong communication skills and the ability to collaborate with highly technical stakeholders.

What This Role Requires

• Genuine hands‑on Machine Learning experience rather than solely software engineering around ML products.
• Strong understanding of modern model training, evaluation, and agent development.
• Ability to operate at Staff level by setting technical direction, not simply executing against predefined requirements.
• Comfort moving between experimentation, infrastructure, architecture, and implementation.
• Ability to influence and lead technically without requiring formal authority.
• A strong builder mindset rather than a purely academic or publication‑oriented background.
• Comfort using AI coding tools as part of day‑to‑day engineering while fully understanding and defending the resulting implementation.

Why Join

• Work on reinforcement learning, agent training, post‑training, evaluation, and other technically challenging areas of modern AI.
• Operate closer to the underlying models and training systems rather than focusing primarily on application‑layer integrations.
• Combine deep ML problems with production engineering responsibility.
• Take meaningful technical ownership while remaining hands‑on.
• Work across a range of fast‑moving AI problems rather than becoming narrowly siloed into a single technical area.
• Join an environment that values speed, high agency, strong engineering judgment, and individual impact.