SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Parachute Health is the leading digital ordering platform for medical equipment and supplies, connecting 300,000+ clinicians and 3,000+ supplier locations across all 50 states. The company has helped 15M+ patients access life-saving products at home.
You'll join an incubator-style AI team scaling AI capabilities across the entire product ecosystem. This is a high-impact individual contributor role where you'll design and build agentic LLM pipelines powering AI features across the product.
Key responsibilities include:
- Design and build agentic LLM pipelines (LangGraph, LangChain, etc.) that power AI features
- Build full-stack product features using Python/FastAPI backends and TypeScript/React frontends
- Prompt-engineer and iterate on LLM prompts and models, evaluating improvements based on accuracy, latency, and cost
- Strengthen evaluation frameworks with curated datasets, evals, experiments, and regression identification
- Improve production observability through accuracy signals and feedback loops
- Contribute to architecture and technical direction
- Partner with product teams to understand AI use cases and enable them effectively
- Advise on when to apply AI/LLM solutions and when to avoid them, championing ROI-driven approaches
Required qualifications:
- 5+ years of professional software engineering experience
- Startup or 0-to-1 environment experience with end-to-end feature ownership
- Strong full-stack web application experience in Python
- Shipped reliable agentic LLM workflows to production
- Experience evaluating and testing LLM workflows (eval harnesses, golden datasets, regressions, observability)
- Familiarity with prompt engineering and iteration lifecycle
- Fluent with AI-powered coding tools (Cursor, Claude Code)
- Strong written and verbal communication skills
- Must reside in the U.S.
Nice-to-have: healthcare/health-tech experience, Ruby on Rails, track record of evaluating and driving AI tool adoption.
