Leads the team this role joins
In-house, owns the opening
Director or head of the group
From your own Gmail and LinkedIn.
About the role
This is an applied AI research role at Midstream, open in the United States on a remote basis, for people with two to eight years of experience in applied AI, machine learning or a related field. Midstream makes an AI-native financial operating system for health systems, and this role focuses on building AI agents and models that find recoverable financial value in contracts, spend and operational data. The researcher works with engineering, product and subject matter experts to turn manual value analysis into repeatable AI workflows, with success measured by the financial impact of the work. The posting says the company is hiring at several levels and values demonstrated ability over years of experience alone.
What you’ll do
- Build, extend and improve core AI agents that find and validate financial opportunities across healthcare supply chain, pharmacy and managed care
- Investigate large, messy healthcare datasets to surface patterns, anomalies and relationships
- Develop models for pricing discrepancies, contract optimization, underpayments and negotiation opportunities
- Work with the Value Engineers to turn their analyses into reliable AI capabilities
- Develop information extraction and agentic workflows, including tooling that speeds up building extractors for complex healthcare documents
- Define evaluation frameworks and metrics, then test and improve models against them
- Partner with engineering and product to shape how healthcare data is modeled on the platform
What they’re looking for
- Two to eight years of experience in data science, machine learning engineering, AI engineering or a related field
- Proficiency in Python with a solid grounding in statistics, experimental design and data modeling
- Hands-on experience building, debugging and improving LLM-powered agents or workflows
- Ability to design evaluations, identify failure modes and improve performance without unnecessary complexity
- Ability to investigate unfamiliar datasets and form well-reasoned hypotheses
- Focus on outcomes, with the ability to tie technical gains to business results
- Comfort with ambiguity, high ownership and clear communication across technical and non-technical teams
What’s on offer
- Remote work in the United States
- High-ownership role where you can test new approaches and see the financial impact of your work directly
Skills
Questions about this role
How much experience does this Midstream role need?
The listing asks for two to eight years of experience in applied AI, machine learning, data science or a related field, and says the company is hiring across several levels.
Is this role remote?
The listing is marked as remote in the United States.
What does the work involve day to day?
You would build and improve AI agents and models that find financial opportunities in health system contracts and spend data, working closely with engineering and product.