LLM Engineer (Healthcare MVP - Short-Term Contract)

Contract Duration: 1 - 2 months (extendable)

Engagement: Full-time contractor / freelancer

Role Overview

You will work closely with a healthcare domain expert to build a functional LLM-powered MVP designed for investor pitching. The goal is to rapidly prototype a healthcare-focused AI product that demonstrates clear technical feasibility, workflow logic, and user-facing value.

This role suits someone who can own the entire development cycle - from clarifying requirements to delivering a working prototype - while communicating progress clearly and making iterative improvements.

Key Responsibilities

• Translate the client’s healthcare domain guidance into a structured MVP plan and technical requirements.
• Build a lightweight but functional LLM-based prototype (chat, workflow automation, data extraction, or clinical reasoning depending on brief).
• Configure and fine-tune models using appropriate frameworks (OpenAI, Anthropic, local models, LangChain/LlamaIndex, etc.).
• Implement basic UX flows or integrate with simple front-end interfaces for demo purposes.
• Ensure the system’s responses align with healthcare accuracy expectations (non-diagnostic).
• Document the architecture, model choices, and demo instructions for investor use.
• Collaborate in short, high-velocity feedback loops with the healthcare expert.

Required Skills & Experience

• 3+ years of experience in AI/ML or software engineering with hands-on experience building LLM applications.
• Strong experience with at least one LLM stack (OpenAI, LangChain, LlamaIndex, RAG pipelines).
• Ability to convert vague domain inputs into a working prototype quickly.
• Experience building MVPs or rapid prototypes.
• Comfortable working independently on a short timeline with minimal supervision.
• Understanding of healthcare workflows or regulated-domain constraints (preferred).

Nice-to-Have

• Experience working in early-stage startups or investor-driven MVPs.
• Basic UI/UX or front-end skills (React, Next.js, Streamlit, Gradio).
• Prior work with structured medical data or ontology mapping (SNOMED, ICD, etc.).

What Success Looks Like

In 30 days, you deliver:

• A functional LLM-based demo that clearly shows the product’s core value.
• A lightweight technical pitch pack (architecture, API diagram, model reasoning).
• A stable prototype the client can use in investor meetings.