Senior Data Scientist

Gifthealth Inc; United States; On-site; USD 140,000 - 190,000; Full time

Job summary

Gifthealth Inc. is seeking a Senior Data Scientist to own traditional machine learning models addressing operational decisions and patient-outcome questions.

You will evaluate agentic AI workflows and collaborate with the AI/ML Engineering team to ensure robust, compliant solutions that scale. The role emphasizes end-to-end modeling—from data sourcing and feature engineering to deployment and monitoring—while mentoring junior team members and contributing to a growing data science practice

Qualifications

• Demonstrated capability building predictive models and integrating them into business processes.
• Experience deploying models in production, including monitoring, retraining, and scoring.
• Strong SQL and Python skills; comfort with cloud warehouses and dbt workflows.

Responsibilities

• Own traditional ML models addressing operational and patient-outcome questions.
• Develop and validate evaluation mechanisms for agentic AI workflows.
• Source and prepare training data with data engineers; manage feature pipelines and end-to-end modeling data.
• Document best practices, monitor models, and mentor newer data scientists as the team grows.

Skills

SQL

Python

Cloud platforms

Model deployment

Data modeling

LLMs evaluation

Documentation

Team mentoring

Tools

Snowflake

Snowflake ML

dbt

Terraform

Looker

Snowpark

SageMaker

Job description

Senior Data Scientist

About Us

At Gifthealth, we're revolutionizing the way people experience healthcare by simplifying the process of managing prescriptions and health services. Our mission is to provide a seamless, personalized, and efficient healthcare experience for all our customers. We're a dynamic, innovative, and customer-centric company dedicated to making a positive impact on people's lives.

Position Summary

We are seeking a Senior Data Scientist to own traditional machine learning models that answer operational and patient-outcome questions, and support the evaluation of agentic AI workflows as our team begins to field generative workflows that act on model scores and real-time signals. This position plays a key role in supporting the AI/ML Engineering team and reports to the Sr. Manager, AI/ML Engineering, ensuring alignment with organizational goals, operational excellence, and compliance standards.This is an individual-contributor role with no direct reports today, though you will mentor newer teammates as the team grows. You'll join a small but growing cross-functional team with high autonomy and minimal ceremony. We are building new services and practices from the ground up, we write things down, and we invest in education, upskilling, and career growth. We deliver quickly and iterate. Data is PHI. HIPAA, BAAs, masking, and role-based access are daily working realities here, not a compliance module you click through once.

Key Responsibilities

• Models against operational decisions: own traditional machine learning models that answer questions such as which patients are likely to miss their next refill, switch pharmacies, or stall before checkout; which new prescriptions will stall in prior authorization; and how to optimize call volume, staffing demand, or distribution center throughput, driving both patient outcomes and operational improvements.
• Agentic AI evaluation: own the evaluation mechanisms that prove agentic AI workflows are reliable, and infuse data-driven rigor into generative workflow development decisions such as foundation model selection and parameter optimization.
• Training data sourcing and preparation: partner with data engineers on the pipelines your features depend on, drive sourcing of new data inputs, and own modeling data end-to-end, within a data platform built on Snowflake, a few hundred dbt models, Terraform, and Looker.
• The data science practice: define how the team evaluates and validates a model, how it's documented, what's monitored, and what gets retired. Document best practices and mentor new data scientists as the team grows.

Qualifications

• Education : A degree is optional; we evaluate demonstrated capability, not credentials.
• Licensure/Certification : Not applicable.
• Experience : Typically 4 to 7 years of experience, as a guide rather than a requirement.
• Knowledge, Skills, and Abilities : You have built predictive models and shipped them into a business process, and you can name the decision each one served.
• You have owned at least part of how your models run after launch: deployment, batch or real-time scoring, monitoring or retraining. Ideally on a cloud ML platform such as Snowflake ML or Amazon SageMaker.
• Strong SQL and strong Python. Comfortable in a cloud warehouse and in dbt, or clearly able to get there fast.
• Real statistical judgment: you know the difference between a model that validates well and a model that will hold up, and you can explain leakage, drift, and a badly chosen target to a non-technical stakeholder.
• You have worked with LLMs seriously enough to be skeptical: you build evaluations, you know where hallucination bites, and you do not confuse a good demo with a working capability.
• Demonstrated application of the above Qualification: You are comfortable being the first: no playbook, no platform, an incomplete warehouse underneath you, and a need to pick the right first problem anyway.
• You can say "this should not be a model" and mean it. Some of what looks like a data science problem here is a broken process or a missing field.
• You exercise strong independent judgment and you mentor newer teammates. Both are level expectations in this role, and you will be the person others bring modeling questions to.
• You use AI tools effectively, but your foundational coding and data science skills are sharp enough that you can read, audit, and debug every line yourself. If an agent writes it, you still own it and can defend it.
• Nice to have: Snowflake, and especially Snowflake ML (Feature Store, Model Registry, Snowpark). This is a strong plus, as it's where our models are built and served today.
• Nice to have: healthcare, pharmacy, payer, or life-sciences data, such as claims, prior authorization, benefit investigation, or adherence measurement.
• Nice to have: generative AI platforms such as Snowflake Cortex AI or Amazon Bedrock.
• Nice to have: working with PHI under HIPAA and BAAs, and the instincts that come with it.
• Nice to have: time ser