Applied AI Researcher

Midstream is an AI-native financial operating system for health systems, founded by experienced entrepreneurs and backed by top-tier investors. We give finance, supply chain, and managed care teams real-time visibility into where margin is at risk, with a precise action plan for capturing it before it's gone.

Purpose-built for healthcare's complexity, Midstream transforms structured, unstructured, and external data into instant, contract-aware insights that drive smarter decisions. Our AI agents unify spend and revenue operations across the full back-office, learning and adapting with every interaction and delivering always-on intelligence that no point solution or one-off engagement can match.

We're compressing months to minutes and turning post-period hindsight into in-the-moment foresight. The tempo of healthcare finance is changing, and Midstream is leading it.

We're looking for Applied Researchers who are excited to push the boundaries of what AI can accomplish in healthcare. You'll develop intelligent agents and models that uncover financial opportunities hidden within complex health system contracts, spend, and operational data, helping some of the country's largest health systems recover meaningful value.

Working closely with Engineering, Product, and our subject matter experts, you'll turn complex, manual analyses into intelligent, scalable workflows. You'll have the opportunity to experiment with new approaches, develop novel capabilities, and directly measure the financial impact of what you build.

This is a high-ownership role for someone who enjoys solving hard problems, moves quickly from experimentation to results, and is motivated by seeing their work make a tangible difference.

Build and improve our core AI agents

Own the development and continuous improvement of agents that identify and validate financial opportunities across healthcare supply chain, pharmacy, and managed care. Experiment with new approaches, systematically improve performance, and measure success by the value your agents uncover.

Turn complex data into actionable intelligence

Investigate large, messy healthcare datasets to uncover patterns, anomalies, and relationships that reveal financial opportunities. Develop models and analytical approaches to problems such as pricing discrepancies, contract optimization, underpayments, and negotiation opportunities.

Automate how we find value

Work closely with our Value Engineers to understand the complex analyses they perform today and translate those workflows into reliable, AI-powered capabilities. Help reduce manual effort while improving the speed, accuracy, and scale of opportunity identification and validation.

Advance our agentic capabilities

Develop novel approaches to information extraction and agentic workflows, including frameworks that make it faster to build and deploy new extractors across complex healthcare documents. Collaborate with Engineering to translate successful research into scalable product capabilities.

Build with rigor and measurable outcomes

Define clear evaluation frameworks and performance metrics from the start. Continuously test, debug, and improve models and agents, ensuring that gains in technical performance translate into measurable financial value for our customers.

Collaborate to shape our platform

Partner with Engineering, Product, and domain experts to improve how we model healthcare data, identify new opportunities for automation, and influence the evolution of our AI-native platform. Bring curiosity, strong technical judgment, and a willingness to learn from the people closest to the problems we're solving.

• Experienced in applied AI and machine learning: You have 2–8 years of experience in data science, machine learning engineering, AI engineering, or related fields. We're hiring across multiple levels and value demonstrated ability over years of experience alone.

• Strong in data analysis and experimentation: You're proficient in Python, with a solid foundation in statistics, experimental design, and data modeling. You can quickly investigate unfamiliar datasets, identify meaningful patterns, and form well-reasoned hypotheses.

• Hands-on with AI agents: You have experience building, debugging, and improving LLM-powered agents or workflows. You understand how to design effective evaluations, identify failure modes, and systematically improve performance without introducing unnecessary complexity.

• A rigorous problem-solver: You combine analytical curiosity with sound technical judgment. You use AI tools to accelerate your work but develop your own understanding of the underlying data, models, and problems you're solving.

• Outcome-oriented: You're motivated by solving real business problems and can connect technical improvements to measurable results. You care as much about the value your work creates as the sophistication of the approach.

• A high-ownership builder: You thrive in fast-moving environments, take initiative when faced with ambiguity, and know when to ask questions, involve others, or independently push a problem toward resolution.

• Collaborative and thoughtful: You actively seek expertise, communicate openly, and work well across technical and non-technical teams. You take ownership of mistakes, welcome feedback, and contribute to a culture of continuous improvement.

• Curious and mission-driven: You enjoy experimenting with new ideas, learning complex domains, and applying emerging technology to meaningful problems. You're excited by the opportunity to use AI to transform how health systems identify and recover financial value.