Software Engineer III, Agent Quality and Efficiency Optimization

Requirements:

• Bachelor’s degree or equivalent practical experience.
• 2 years of experience with software development in one or more programming languages such as Go, C++ or Python, or 1 year of experience with an advanced degree.
• 2 years of experience with machine learning, large language models (LLMs), or applied AI.

Nice to haves:

• Master's degree or PhD in Computer Science, Machine Learning, or a related technical field.
• Experience building, deploying, or optimizing LLM applications or agentic frameworks (e.g., Agent Development Kit, LangGraph, LangChain).
• Experience with one or more of the following: model serving optimization, inference caching, model routing or LLM cost and latency profiling.
• Experience instrumenting, benchmarking, or evaluating ML systems in production.
• Experience with technical leadership, mentoring engineers, or driving cross-team engineering initiatives.

What you'll be doing:

• Design, build, and ship agent efficiency systems, per-task cost and quality telemetry, smart model routing, durable and deferred execution, and continuous optimization, into the Agent Development Kit and Vertex AI serving stack.
• Measure where cost and latency go, and turn that data into targeted engineering work.
• Shape the team's efficiency roadmap, and help set technical direction for how enterprise agents run in production.
• Benchmark efficiency techniques on both open source benchmarks and real first-party and enterprise agent workloads to prove cost and latency wins hold up without regressing accuracy.
• Partner with the team's evaluation engineers, model-serving teams, and product teams to close optimization loops on measured behaviour and land changes across Agent Platform.

Perks and benefits:

• Individual pay is determined by factors including job-related skills, experience, and relevant education or training.
• Poland: zł280000 - zł287000 (PLN) + 15% bonus target + equity + benefits