4 months ago
Responsibilities
- Lead the design and architecture of agentic AI systems with reasoning loops, tool calling, dynamic multi-agent orchestration, RAG pipelines, memory and state management, MCP, and A2A.
- Build and own production agent infrastructure including prompts, function tools, workflow graphs, protocol integrations, and adaptive agent lifecycle management.
- Develop evaluation, testing, benchmarking, regression, adversarial testing, safety guardrail, observability, and human-in-the-loop frameworks.
- Train, fine-tune, and deploy large-scale machine learning models on healthcare datasets in collaboration with researchers and clinicians.
- Shape the company’s agentic AI strategy and roadmap with the C-suite and cross-functional leadership.
- Prototype and iterate on novel agent behaviors and help establish transparent, compliant healthcare AI systems.
Requirements
- 7+ years of hands-on Machine Learning Engineer experience with production agentic AI systems.
- Experience designing, optimizing, or integrating analytic engines or advanced analytics platforms.
- Strong theoretical foundation in machine learning and AI, especially NLP, LLMs, reinforcement learning, and planning or reasoning algorithms.
- Expertise with LangChain, LangGraph, Model Context Protocol, Agent-to-Agent protocols, Hugging Face, PyTorch, vector databases or semantic search, prompt engineering, and observability platforms such as LangSmith and Phoenix.
- Experience building automated evaluation and testing pipelines for autonomous agents, including reliability, safety, factuality, cost, latency, clinical utility, and dynamic behavior metrics.
- Preferred qualifications include healthcare data experience with EHRs or claims, knowledge of HIPAA and related regulatory considerations, top AI/ML conference or journal publications, and AWS, Azure, or GCP experience.
Benefits
- Ownership from day one in a small, high-trust team
- Opportunity to build AI-powered healthcare workflows with real-world impact
- Series A company backed by top-tier investors with access to 200M+ patient records
Tech Stack
Categories
About Medeloop
Medeloop is an agentic platform that brings intelligence to healthcare data, wherever it lives. Medeloop originated from Stanford University and is backed by General Catalyst, Icon Ventures, Inovia Capital, Healthier Capital, and more. Medeloop's team includes leaders in AI, life sciences, and medical research.