Success in this Staff Applied Scientist role is defined by metrics that bridge algorithmic performance with tangible healthcare outcomes. Key technical indicators include:

• Algorithmic Efficacy: Measuring improvement in provider tiering optimization (balancing geographic access versus cost-of-care savings) and the precision/recall of member engagement models.
• Operational Impact: Quantifying the reduction in healthcare costs and improvements in member health outcomes directly resulting from model-driven interventions.
• Safety & Reliability: For the AI primary care tool, metrics focus on robust evaluation harnesses, hallucination rates, and medical safety guardrails.
• Scalability: The ability of shipped systems to maintain performance across a dataset of 320M+ patients while consistently driving measurable changes in physician selection and patient behavior.

At Garner, I balance ML and heuristic approaches by prioritizing the problem’s objective over technical complexity. I follow a "fit-for-purpose" philosophy: if a transparent, rule-based heuristic or optimization model achieves the business goal with higher reliability and interpretability, I choose it. Conversely, I deploy advanced ML or LLMs when the problem space—such as member engagement or LLM-based primary care—demands the nuance and pattern recognition only high-dimensional models provide. In production, I define rigorous validation harnesses and clear metrics early on. This allows me to objectively measure trade-offs between precision and simplicity, ensuring that every algorithmic choice is sound, scalable, and directly drives improvements in healthcare outcomes for our members.

At Garner, the Applied Science team is adapting to the evolving LLM landscape by prioritizing safety, rigor, and outcome-driven frameworks. Rather than relying solely on off-the-shelf metrics, they are building custom evaluation harnesses and guardrails specifically designed for medical-adjacent products. This ensures that LLM-based systems, such as their "AI primary care doctor," are both reliable and safe for production. The team emphasizes a practical, tool-agnostic approach—leveraging modern LLM tooling while choosing the right methodology for the problem rather than just chasing hype. By integrating these evaluations into the end-to-end development cycle, they maintain high standards for technical accuracy and patient trust while quickly translating prototypes into high-impact, real-world solutions.

The team prioritizes between provider tiering and member engagement by aligning technical approaches with specific business outcomes. Both are treated as high-stakes, end-to-end production problems where success is measured by rigorous, predefined metrics. Rather than favoring one over the other, the team evaluates the unique constraints and objectives of each: provider tiering focuses on optimizing geographic access and cost-of-care, while member engagement utilizes claims and behavioral data to optimize intervention timing and channels. The Staff Applied Scientist acts as the primary decision-maker, framing these ambiguous problems into clear, actionable frameworks and selecting the appropriate methodology—whether ML, optimization, or heuristics—to ensure both projects move the company’s most critical performance indicators.

Garner’s product strategy centers on a "win-win" value proposition that aligns clinical excellence with financial efficiency. By applying 550+ proprietary clinical metrics to a massive dataset of 320M+ patient records, the platform identifies high-performing doctors who provide superior care. The company then leverages data-driven incentives to steer members toward these providers.

This approach balances quality and business goals by treating healthcare delivery as an optimization problem: improving health outcomes naturally reduces total costs. Garner utilizes algorithmic systems—ranging from provider tiering to member engagement models—to navigate these complex trade-offs. Ultimately, the strategy ensures that the highest-quality care path is also the most cost-effective, creating scalable, measurable impact for both employers and patients.