SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Garner Health is transforming the U.S. healthcare system by partnering with employers to redesign healthcare delivery. The company applies 550+ proprietary clinical metrics across 80+ specialties to a dataset of 320M+ patients to identify top-performing doctors and steer members toward higher-quality, lower-cost care. In five years, Garner has helped 2.5 million people access better care and saved $1B in healthcare costs.
As an Applied Scientist II, you will own production algorithmic systems end-to-end—from problem framing and objective definition through shipping and real-world iteration. This is not a dashboards or descriptive analytics role; the closest analog is a quantitative researcher at a top hedge fund. You will design and ship the algorithmic systems at the core of Garner's product, answering critical questions like "Which doctor should I see?" and "What will it cost?"
Key responsibilities include:
- Ship production algorithmic systems end-to-end, from problem framing to launch to iteration
- Frame messy healthcare and business constraints into clear objectives, tradeoffs, and decision frameworks
- Define metrics to judge solution effectiveness and validate before shipping
- Choose the right approach for each problem: machine learning, optimization, heuristics, expert systems, or hybrids
- Experiment with novel approaches to achieve better results
- Produce rigorous, well-validated work and review peers' code and analyses
- Build deep understanding of the healthcare economy and Garner's role in it
Near-term roadmap problems include: (1) Provider tiering optimization—building an algorithm that jointly optimizes geographic access and total-cost-of-care savings; (2) AI primary care doctor—fine-tuning and productionizing an LLM-based primary care experience with evaluation harnesses and quality monitoring; (3) Member engagement model—building an ML system that chooses the right channel and moment for touchpoints (SMS, push, phone, email) to influence member behavior.
You will work with Python, SQL, AWS, Snowflake, pandas, XGBoost, PyTorch, HuggingFace, and modern LLM tooling. The role is based in the NYC Financial District office, requiring 3 days per week in-office (Tuesday, Wednesday, Thursday).
