SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Abridge is an AI-powered healthcare platform that transforms patient-clinician conversations into structured clinical notes in real-time. Founded in 2018, the company has built enterprise-grade technology with deep EMR integrations and is a pioneer in responsible generative AI deployment across health systems.
As a GenAI Software Engineer, you will design and build agentic systems that turn large language models into composable, dependable clinical tools. Your work will span the full lifecycle of LLM-powered workflows: architecting systems that leverage retrieval, tool use, agentic reasoning, and structured outputs; collaborating with ML and infrastructure engineers to optimize for latency and uptime; and owning end-to-end productionization including deployment, monitoring, observability, and guardrails.
A major focus is developing rigorous evaluation frameworks—both automated and human-in-the-loop—to continuously assess accuracy, robustness, and multilingual capabilities. You'll design experiments to probe failure modes, stress-test workflows under realistic conditions, conduct A/B tests against baselines, and translate clinician feedback into actionable model improvements.
You'll write high-quality, modular Python code that gracels under failure and is easy to iterate on. You'll own major architectural decisions around workflow design, data flow, caching, and output structuring. You'll also rapidly prototype with frontier models, open-source tools, and novel prompting techniques.
The role requires 3+ years building production systems, with 1–2+ years focused on LLM or agentic products. You should have deep fluency with LLM APIs, prompting strategies, orchestration patterns (LangChain, LlamaIndex), retrieval systems, function calling, and tool-use. Strong Python fundamentals, experience with async programming and deployment tooling, and familiarity with model evaluation and A/B testing are essential. This is a hybrid role requiring 3 days per week in the SF or NYC office.
