Intern – Machine Learning & Generative AI

Summary

Join EvenUp, a rapidly growing generative AI startup, and contribute to our mission of leveling the playing field for personal injury victims. We are developing cutting-edge document intelligence and conversation automation, using large language models to automate tasks in legal and medical workflows. As a research intern, you will research and prototype new approaches in generative AI, fine-tune models for legal and medical corpora, build autonomous systems, and create evaluation suites for safety and bias. You will work closely with a team of engineers, AI researchers, and legal experts, and have the opportunity to publish your work. EvenUp offers a fast-paced, applied research environment where your contributions will have a direct impact on real-world cases.

Requirements

• Pursuing a Ph.D. in CS, ECE, Statistics, or related field with a research focus in ML / NLP / Generative AI
• Strong grasp of modern LLM architectures (transformers, attention variants), training/fine-tuning pipelines (LoRA, PEFT, RL-HF), and evaluation methods
• Fluency in Python
• Demonstrated ability to turn research ideas into working prototypes (open-source projects, or industry experience)
• Intellectual curiosity, rapid learning loop, and the grit to thrive in an ambiguous, 0-to-1 environment

Responsibilities

• Research & Prototype Design and evaluate new approaches in retrieval-augmented generation, tool-use agents, multimodal LLMs, or self-supervised document representation—whichever unlocks the next pain-point for our users
• Domain Adaptation Fine-tune and align models to the quirks of long, jargon-dense medical/legal corpora (ICD codes, CPTs, deposition transcripts, insurance clauses)
• Agentic Systems Build and benchmark autonomous chains that decide when to call OCR, which precedents to cite, or how to interview a claimant—all with safety and auditability baked in
• Measurement & Safety Create eval suites for factuality, legal soundness, and bias. Propose mitigations where gaps appear
• Publish & Share If your work advances the field, we’ll support conference submissions—credit where credit is due

Preferred Qualifications

• Experience with agent frameworks (e.g., LangGraph, AutoGen) or building tool-using LLM agents from scratch
• Hands-on with long-context or multimodal models
• Familiarity with vector databases, RAG orchestration, or document triage pipelines (OCR, layout parsing)

Benefits

• Impact on Day 1 – Your code runs on live cases, not in a sandbox
• Mentorship – Weekly 1-on-1s with our Head of AI and access to attorneys for domain deep-dives
• Publication Support – Travel stipend and legal review for papers or posters
• Conversion Path – High-performing interns are first-in-line for full-time research scientist or ML engineer offers
• Choice of medical, dental, and vision insurance plans for you and your family
• Flexible paid time off
• 10 US observed holidays, and Canadian statutory holidays by province
• A home office stipend
• 401(k) for US-based employees
• Paid parental leave
• Sabbatical program
• A meet-up program to get together in person with colleagues in your area
• Offices in San Francisco and Toronto