About Healthcare.com
Healthcare.com is a healthcare technology company building smarter ways to connect people and partners with the coverage, care, and solutions they need.
We’re three businesses, one company. Through Marketplace, Pivot Health, and TrustRx, our work spans health insurance marketplaces, flexible health coverage, and pharmacy, giving us more ways to solve meaningful problems across the healthcare ecosystem.
Healthcare doesn’t stand still, and neither do we. We’re curious, collaborative, and always looking for better ways to solve problems, use technology, and make an impact.
The Opportunity
Healthcare.com is putting AI at the center of how our data platform serves the business. We already have a modern warehouse, strong pipelines, and a small, highly leveraged senior team. What we need now is someone to own the layer on top: the retrieval systems, agentic workflows, and self-service AI tooling that let people across the company get answers and take action without waiting in a queue.
This is a senior, hands-on individual contributor role reporting directly to the VP of Data. You will work shoulder-to-shoulder with our senior data engineers and data scientists, and you will be the person business teams come to when they want something built with AI. You will reduce the load on our senior engineers by independently shipping production systems — not by adding coordination overhead.
This is a pragmatic builder role. Success looks like useful systems running in production under messy real-world conditions, a higher architectural bar across the team, and stakeholders who are unblocked. It does not look like papers, benchmarks, or experiments that never ship.
Where You'll Make an Impact
Own AI systems end to end — from ingestion and transformation through serving and the AI layer on top. You will not be handed a spec and a model endpoint; you will design the whole path.
Build production retrieval systems on our data — RAG and hybrid/semantic search grounded in our warehouse, with the ingestion, chunking, embedding, and freshness work that makes retrieval actually trustworthy.
Design and ship agentic workflows — tool-using assistants and multi-agent systems with real production concerns handled: tool and API integration, context management, session state, memory, caching, and failure behavior.
Stand up a self-service AI and analytics layer serving 200+ internal stakeholders across marketing, sales, operations, product, and finance — with governance and access controls built in rather than bolted on.
Make AI quality measurable — build the eval frameworks, tracing, observability, and monitoring that tell us when a system regresses, plus the guardrails and cost controls that keep it viable at scale.
Raise the architectural bar — establish reusable patterns, shared libraries, deployment templates, and CI/CD so the next AI system takes a fraction of the effort of the first.
Partner directly with business teams — translate vague business problems into scoped technical work, and push back credibly when a requested AI feature will not work or will not scale.
Level up the people around you — mentor engineers, set technical direction, and bring the rest of the team along on AI patterns, without carrying a management chain.
What Success Looks Like
You are productive within your first 90 days without consuming large amounts of senior engineering time, and you have shipped at least one useful AI workflow into the hands of real users.
By six months, at least one AI system you own is running in production with evals, monitoring, and cost controls in place — and there is a documented reference architecture others can build against.
Business stakeholders route AI and self-service requests to a working system instead of to a person's backlog.
What You'll Bring to the Team
7+ years of hands-on engineering experience across data and software — you have built and operated real systems, not just analyzed them.
2+ years shipping production LLM/AI systems: RAG, agents, evals, prompt orchestration, or model-integrated pipelines. Prototypes that stayed prototypes do not count; we want to hear about something users depend on.
Deep foundation in data engineering or applied data science — strong Python and SQL, modern cloud data warehousing (Snowflake, Databricks, BigQuery, or similar), and orchestration (Airflow, dbt, or similar).
Cloud engineering experience, AWS preferred — comfortable deploying containerized and serverless services, and building the CI/CD around them.
A track record of designing systems end to end, where you owned the architecture and the trade-offs, not just a component of someone else's design.
Demonstrated autonomy in ambiguous environments — you can scope, design, and ship without detailed product specs or close oversight, and you know when to come back and ask.
Strong systems thinking — you care about architectural consistency, governance, evals, observability, and cost, not only about shipping the first thing that works.
Clear communication with both technical peers and non-technical stakeholders, including senior leadership. You can explain a trade-off to an executive and a constraint to a marketer in the same afternoon.
Authorization to work in the United States without visa sponsorship, now and in the future.
What Sets You Apart
Experience in healthcare, insurance, fintech, or another regulated domain working with sensitive data (PHI, PII) and the compliance constraints that come with it.
Knowledge graphs and graph databases (Neo4j, Cypher) — especially graph algorithms applied to pattern detection, entity resolution, or feature extraction.
Model Context Protocol (MCP) servers or comparable tool-integration layers that let agents reach warehouses, APIs, and internal systems safely.
Modern agent and LLM tooling — LangChain, CrewAI, Strands, Bedrock AgentCore, Anthropic/OpenAI APIs — and LLM ops tooling such as Langfuse for tracing and evaluation.
Vector search in production (pgvector, Pinecone, Weaviate, or warehouse-native alternatives).
Infrastructure and delivery depth — Docker, Kubernetes, Terraform, Jenkins or GitHub Actions, package publishing and versioning.
Building lightweight internal tools and UIs (Streamlit, Next.js/React) to put AI capability directly in stakeholders' hands.
Experience leading a small engineering team or setting technical direction as a senior IC.
Bachelor's degree in computer science, information systems, engineering, or a related field — or equivalent practical experience.
What This Role Is Not
A research role. We are not staffing paper writing or open-ended experimentation.
A people-management role. You will mentor and set direction, but you will not have direct reports.
A prompt-tuning role. The hard parts here are data, retrieval quality, integration, evaluation, and operations.
