AI Engineer - Quality and Post-Training

Description

The company builds an enterprise AI platform with search, an AI assistant and AI agents, offering 100+ SaaS connectors, a choice of LLMs and open APIs so organizations can run AI without vendor lock-in.

Its knowledge graphs map links between people, content and activity so answers suit each employee, and its agents carry out multi-step work using internal and public data, both historical and live.

The company has 1,000+ staff and serves customers in 50+ industries across more than 25 countries.

The role sits between applied research, quality and product engineering. The main challenge is measuring what good output looks like and closing the gap, more than writing code.

The focus is raising quality across models and agent harnesses: running evals, improving evalsets and verifiers, post-training models for routing and enterprise tasks, and bringing new research into the product with product teams and customers.

Responsibilities

• Use eval results and production traces to improve the whole stack, including the agent harness, model routing, prompts and retrieval, balancing quality against latency and cost.
• Post-train models using supervised fine-tuning and preference or reward-based methods such as DPO, PPO and GRPO, including producing the training data.
• Create RL environments that simulate enterprise workflows, with evalsets, harness support and LLM-based verifiers used for hill-climbing and for internal and external benchmarks.
• Refine representative evalsets, LLM judges and agentic verifiers, and calibrate them against human graders so scores hold up in release and training decisions.
• Partner with product, platform and research teams to get quality gains into production.

Requirements

• 5+ years of industry experience in AI or machine learning engineering.
• Hands-on experience post-training LLMs, including adapting open-weight models to specialized tasks.
• Experience building evals for assistant and agent systems: generating evalsets, LLM-as-a-judge verifiers, and calibrating offline and online signals. You know when a metric truly reflects better quality.
• Strong coding in Python and/or Go, with TypeScript a plus. You are proficient in your preferred ML framework and have shipped production ML systems.
• Product-minded: you turn vague problems into concrete experiments and do well in a close, cross-functional team.

Conditions

Hybrid: 4 days a week in the Mountain View or San Francisco office.

Base salary of $200,000 to $280,000 per year. Final pay depends on location, level, job-related knowledge, skills and experience. Some roles may also be eligible for variable pay, equity and benefits.

The interview includes a short AI-focused exercise or discussion on how you think about, design and use AI. Prior experience with the company is not required, and you may reference any tools or workflows you already use.

The company is committed to a diverse, inclusive workplace and does not discriminate on the basis of gender, ethnicity, sexual orientation, religion, civil or family status, age, disability or race.

Benefits

• Competitive compensation.
• Medical, vision and dental coverage.
• Generous time-off policy.
• 401k contributions to support long-term savings.
• Home office improvement stipend when you join.
• Annual education and wellness stipends.
• Regular company events and healthy lunches every day.