Job Title: AI/NLP Engineer – LLM Prototype for Ethics + Research Compliance (SBIR Phase I) - Contract to Hire

Qualifications

• A fine-tuned or adapter-based LLM (e.g., using GPT, Mistral, LLaMA)
• A working retrieval-augmented generation (RAG) prototype using LangChain, LlamaIndex, or similar
• NLP pipelines for tasks like classification, logic chaining, and policy-aware reasoning
• Fine-tuning large language models
• Building RAG systems and vector databases
• Experience with compliance-heavy domains (e.g., healthcare, law, finance, govtech)
• Ability to collaborate cross-functionally to scope and refine an idea into a technical MVP
• Bonus: Previous NIH or grant-backed tech work, explainability tools, or UX-aware development

Benefits

• 📅 Timeline & Commitment
• Start date: Early 2026 (if funded)

Responsibilities

• The tool will use custom large language models (LLMs) to help IRB analysts and researchers navigate ethically and legally complex documents like research protocols and informed consent forms
• Our goal is to build a domain-specific GPT-style assistant that can analyze, structure, and respond to dense, regulatory content
• We’re starting from a validated concept and need your help to turn that idea into a functional MVP that demonstrates technical feasibility
• A proof-of-concept model that runs on real IRB materials and showcases the tool’s potential for reviewers

Full Description

We’re looking for an AI/NLP Engineer to join our NIH SBIR Phase I grant proposal for a smart compliance assistant designed to support research oversight teams. The tool will use custom large language models (LLMs) to help IRB analysts and researchers navigate ethically and legally complex documents like research protocols and informed consent forms.

Our goal is to build a domain-specific GPT-style assistant that can analyze, structure, and respond to dense, regulatory content. We’re starting from a validated concept and need your help to turn that idea into a functional MVP that demonstrates technical feasibility.

Note: This project is contingent on funding through an NIH SBIR Phase I grant.

Important: At this stage, I’m seeking team members who are willing to be listed in the proposal. If you’re interested and qualified, all I’ll need is a brief bio or NIH-formatted biosketch.

🔧 What You'll Build

A fine-tuned or adapter-based LLM (e.g., using GPT, Mistral, LLaMA)

A working retrieval-augmented generation (RAG) prototype using LangChain, LlamaIndex, or similar

NLP pipelines for tasks like classification, logic chaining, and policy-aware reasoning

A proof-of-concept model that runs on real IRB materials and showcases the tool’s potential for reviewers

✅ Ideal Skills & Experience

Fine-tuning large language models

Building RAG systems and vector databases

Experience with compliance-heavy domains (e.g., healthcare, law, finance, govtech)

Ability to collaborate cross-functionally to scope and refine an idea into a technical MVP

Bonus: Previous NIH or grant-backed tech work, explainability tools, or UX-aware development

📅 Timeline & Commitment

This project is contingent on NIH SBIR Phase I funding

Start date: Early 2026 (if funded)

Estimated commitment: ~600 hours during Phase I

Why Join?

You’ll get to shape the foundation of a novel AI tool tackling one of the most ethically important and technically underserved areas in research. We value engineers who think critically, communicate clearly, and care deeply about responsible AI.