Technical Staff Member

This is a fully remote position, open to applicants in United States, +1 more country.

📋 Description

• Take charge of the agentic system that translates members’ health data into actionable next steps in collaboration with clinicians.

• Design and manage the complete lifecycle of production multi-agent systems, encompassing orchestration graphs, tool utilization, memory, retrieval, evaluations, and monitoring.

• Integrate cutting-edge LLMs with Function’s exclusive health data.

• Ensure safety, latency, and cost efficiency in healthcare AI applications.

• Design and construct stateful, graph-based agent workflows that incorporate tool usage, planning, and memory.

• Combine LLMs and multimodal models through structured input/output and function/tool invocation.

• Develop highly reliable APIs and streaming services for real-time inference, speech processing, and vision applications.

• Take ownership of production readiness by implementing tracing, logging, metrics, rate limiting, circuit breakers, and service level objectives (SLOs).

• Create evaluation pipelines featuring golden sets, utilizing LLMs as judges, human rubrics, online A/B testing, and continuous integration regression tests.

• Implement hybrid search techniques, vector and graph retrieval, semantic caching, and long-context handling.

• Enhance model routing, prompts, tools, quantization, and key-value caching/prefill strategies to optimize cost and latency.

• Collaborate across functions to convert research findings into robust production systems.

• Guide engineers through the development of design documents and architecture decisions.

⛳️ Requirements

• Minimum of 1 year of experience in building agentic AI systems.

• At least 6 years of experience as a full-stack or machine learning engineer developing production backends or ML systems using Python, Go, or similar languages.

• Proficient in agentic orchestration tools, such as LangGraph, PydanticAI, DSPy, or LlamaIndex, along with function/tool invocation.

• Experience in integrating advanced LLMs and multimodal models through managed APIs or self-hosted services.

• Strong background in API design and backend frameworks, particularly with FastAPI or Flask.

• Familiarity with event-driven architecture.

• Expertise in data systems, particularly PostgreSQL, including token streaming and throughput optimization.

• Experience with vector databases like pgvector, Pinecone, Weaviate, or Milvus, as well as hybrid search and graph/knowledge storage.

• Experience with production evaluations, including LLMs as judges, human-in-the-loop processes, rubric design, and CI-integrated regression testing.

• Knowledge of observability and site reliability engineering (SRE) practices utilizing OpenTelemetry traces, metrics, structured logs, SLOs, dashboards, and on-call triage.

• Cloud-native delivery experience with Kubernetes, Terraform, Docker, and GPU scheduling/autoscaling on AWS or GCP.

• Proficient in CI/CD practices using GitHub Actions and automated testing for prompts, tools, and agents.

• Strong communication skills and a sense of ownership in fast-paced environments.

• Preferred experience with real-time multimodal systems, retrieval-augmented generation (RAG), safety and governance, regulated data, personalization, long-term memory agents, consumer-scale AI applications, high-traffic systems, or on-device/edge acceleration.

🏝️ Benefits

• Competitive salary and comprehensive benefits package.

• Flexible working hours.

• A dynamic work environment that fosters creativity and innovation.