AI that actually understands healthcare is transforming operations, one workflow at a time.
The Role
Build and optimize production AI pipelines for healthcare workflows using LLMs, classical ML, VLMs, OCR, retrieval, agents, and fine-tuned models. Develop document extraction systems, evaluation harnesses, monitoring, and inference orchestration. Improve model quality, latency, cost, explainability, and safety while collaborating with product, engineering, senior ML engineers, and healthcare domain experts.
Summary Generated by Built In
About Autonomize AI
The Opportunity
What You'll Do
Qualifications
Nice to Have
Who You Are
What We Offer
How to Apply
Autonomize AI is an enterprise intelligence platform transforming how healthcare organizations access, orchestrate, and act on complex operational and clinical knowledge. Recognized as a World Economic Forum Technology Pioneer 2026, a Top 100 company on the Inc. 5000, and trusted by three of the five largest U.S. health enterprises, Autonomize AI combines healthcare-specific AI agents, workflow intelligence, and deep domain expertise to enable smarter decision-making across utilization management, care management, claims, pharmacy, and appeals. Backed by Valtruis, Asset Management Ventures, Cigna Group Ventures, ATX Venture Partners, and others. We're building the critical infrastructure for AI-native healthcare operations. We're growing fast and looking for bold, driven teammates to join us.
We're looking for a hands-on AI Engineer who's excited to ship real-world applications, not just benchmarks. You'll build and optimize the AI-native systems behind our healthcare agents and copilots, blending LLMs, vision models, and classical ML across structured and unstructured data to power decisions in high-stakes workflows. This role is for someone ready to go deep on applied problems, from retrieval to routing to generation, and push solutions into production fast.
- Build production AI pipelines. Own and optimize pipelines that combine LLMs and classical ML (RAG, extraction, scoring, summarization, classification) for utilization management, payment integrity, claims, and appeals.
- Extract from real-world documents. Implement VLM- and OCR-based pipelines that turn medical documents, faxes, and healthcare forms into structured, reliable data.
- Fine-tune and adapt models. Run SFT and parameter-efficient (e.g., LoRA) fine-tuning experiments on open-source and proprietary models.
- Implement retrieval and agents. Build retrieval strategies, prompt chains, tool-using agents, and inference orchestration (e.g., LangGraph) for production use cases.
- Build evals and analyze errors. Create eval harnesses and test sets, run error analysis, and turn findings into measurable accuracy gains.
- Monitor and improve. Track and improve model quality, latency, cost, explainability, and safety in production.
- Experiment fast. Prototype new techniques from recent research and help decide what's ready for production.
- Collaborate. Work closely with senior MLEs, product, engineering, and domain experts, and document your work clearly.
- 2+ years experience working in applied ML and LLMs.
- Strong Python skills and familiarity with PyTorch, Hugging Face Transformers, and LLM frameworks (LangChain, LangGraph, LlamaIndex).
- Comfort with embeddings, vector search, retrieval pipelines, and prompt engineering.
- Experience fine-tuning or adapting models, plus working knowledge of classical ML and NLP.
- Understanding of model evaluation, observability, and responsible AI practices.
- Experience deploying models to production and familiarity with MLOps tooling (MLflow, Docker, Kubernetes).
- Solid software engineering fundamentals and clean, testable code.
- A bias for experimentation, clarity, and shipping fast.
- Experience with healthcare, compliance-sensitive data, or regulated environments.
- A BS/MS in Computer Science, Engineering, Data Science, or a related field, or equivalent experience.
- Experience with VLMs or document AI.
- Exposure to healthcare payer workflows (UM, claims, prior authorization, medical coding).
- Experience with agent frameworks or multi-step reasoning systems.
- Open-source contributions, side projects, or technical writing.
- Owner mentality. The buck stops with you. You own it, you learn it, and you get it done.
- Naturally curious. You'd rather experiment than hypothesize. You push boundaries and experiment your way through any problem.
- Passionate and unafraid. You're committed to the team and the mission.
- A team player. You love to learn and win together.
- A strong communicator. You're clear in voice, writing, chat, and video.
- Real-world impact. Your models run in production at some of the largest U.S. health enterprises, shaping decisions in prior authorization, claims, and appeals.
- Category-defining AI products. Help build the AI operating layer for healthcare, a space where the playbook is still being written.
- Hard, unsolved ML problems. Work on VLM document understanding over messy real-world faxes, agentic reasoning over clinical policy, domain fine-tuning (SFT, LoRA, RL), and models that learn from expert feedback in production.
- Research to production, fast. Read the paper, prototype it, ship it. New ideas reach customers in weeks, not quarters.
- Significant ownership and autonomy. Help shape AI strategy, not just execute it.
- Modern stack and compute. You'll have GPU resources and access to the latest open-source and frontier models.
- Build your public profile. Publish papers, file patents, write for our tech blog, and speak at conferences.
- Learn with strong peers. Grow through design reviews, paper discussions, and close collaboration with ML engineers and clinical experts.
- A high-growth environment with exceptional technical challenges.
- Competitive compensation with performance incentives.
- 100% employer-paid health, vision, and dental insurance.
- Retirement plans (401k), disability insurance, and employee assistance programs.
Send your resume and a brief cover letter to [email protected] explaining why you're the right partner for this mission.
Skills Required
- 2+ years of experience working in applied machine learning and large language models
- Strong Python skills
- Familiarity with PyTorch
- Familiarity with Hugging Face Transformers
- Familiarity with LLM frameworks such as LangChain, LangGraph, or LlamaIndex
- Knowledge of embeddings, vector search, retrieval pipelines, and prompt engineering
- Experience fine-tuning or adapting models
- Working knowledge of classical machine learning and natural language processing
- Understanding of model evaluation, observability, and responsible AI practices
- Experience deploying models to production
- Familiarity with MLOps tooling such as MLflow, Docker, and Kubernetes
- Solid software engineering fundamentals and ability to write clean, testable code
- Experience with healthcare, compliance-sensitive data, or regulated environments
- BS or MS in Computer Science, Engineering, Data Science, or a related field, or equivalent experience
- Experience with vision-language models or document AI
- Exposure to healthcare payer workflows, including utilization management, claims, prior authorization, or medical coding
- Experience with agent frameworks or multi-step reasoning systems
- Open-source contributions, side projects, or technical writing
Am I A Good Fit?
Get Personalized Job Insights.
Our AI-powered fit analysis compares your resume with a job listing so you know if your skills & experience align.
Success! Refresh the page to see how your skills align with this role.
The Company
What We Do
Autonomize AI Agents & Copilots organize, contextualize and summarize unstructured data to reduce the administrative burden for healthcare knowledge workers to make data-driven decisions and improve patient outcomes. Our customers include health plans, providers and life sciences companies. Unlike generic AI systems retrofitted for healthcare, Autonomize deeply understands medical contexts, terminologies, and operational nuances. Our healthcare-focused AI Agents & Copilots augment knowledge work, drastically reducing administrative burden. Care management teams spend 78% less time per case, achieving an impressive 85% boost in case review efficiency. Prior authorization processes that traditionally take 20-30 minutes shrink to mere seconds, accompanied by an 80% reduction in manual errors, saving millions of dollars annually. Our AI Agents turn chaotic, unstructured healthcare data—clinical notes, PDFs, faxes, and claims—into structured, contextual information that informs decisions and actions. This has driven substantial real-world impact: organizations using Autonomize experience a 92% reduction in manual effort for care gaps and HEDIS chart reviews, dramatically improving compliance and STAR ratings. Autonomize AI is purpose-built for healthcare, transforming healthcare operations one workflow at a time through AI-native solutions that deliver immediate, scalable impact.
