Director, Engineering (Patient)

Headway's mission is to build a new mental health care system that everyone can access. We've built technology that takes the hardest parts of mental healthcare finding the right provider, navigating insurance, managing payments — and makes them simple. We're now one of the fastest-growing companies in healthcare, with more than 200,000 patients finding care through Headway and thousands of therapists using our platform to grow their practices.

We believe AI is the unlock for our next phase — more intelligent matching, more personalized patient experiences, and a fundamentally different way of building software. This role sits at the center of that transition.

The Patient Core Experience team is building what we believe will become one of the most consequential AI products in healthcare: a system that connects patients with therapists who are genuinely right for them.

Our four engineering pods — Onboarding, Profiles + Checkout, Ranking + Relevance, and Activation — own the full patient journey from first visit through early retention. Today, our matching is largely filter-based. We're rebuilding it as an intelligent system that uses communication style, data-backed expertise signals, patient-reported outcomes, and behavioral signals to surface the right provider for each patient at the moment they're ready to get help.

We operate a three-sided marketplace (patients, providers, payers) with real complexity and massive impact. The leap from directory to intelligent matching platform is the defining technical challenge for this team. It requires both serious AI/ML product work and a leader who has already made the shift to AI-augmented engineering — and has opinions about what that means in practice.

• Matching quality is the foundation of therapeutic outcomes — we obsess over getting patients to the right provider
• AI is how we get there — from ML-powered ranking to LLM-assisted patient experiences to AI-augmented engineering workflows
• We ship fast and learn fast, but patient trust comes first — especially when AI is influencing decisions about someone's mental healthcare
• We build for the whole marketplace: patients, providers, and payers
• Explainability matters — in healthcare, "the model said so" isn't enough

This role exists because we have hard, specific challenges that need senior engineering leadership to crack.

Build the AI-powered matching engine that defines Headway's next chapter. Our current matching is filter-based. You'll own the technical strategy and execution for moving to ML-powered ranking — incorporating provider communication style, clinical expertise signals, and patient outcome data. This means deciding when to use ML models vs. heuristics, how to reason about explainability and bias in a healthcare context, how to A/B test matching quality without degrading patient experience, and how to build patient trust in AI-driven recommendations. Getting this right will directly improve outcomes for every patient on Headway.

Ship AI product features that make the patient journey feel intelligent. Matching is just the start. There's significant opportunity to use LLMs and generative AI to improve how patients understand their options, how we guide them through onboarding and intake, and how we keep them engaged through the early sessions where drop-off risk is highest. You'll own the AI product strategy for your pods — where we go beyond ML ranking into generative and agentic approaches, and how we do it responsibly in a regulated healthcare environment.

Define what AI-era engineering looks like for your teams. You'll set concrete standards for how your ~30 engineers use AI in their workflow — and this isn't a generic "adopt Cursor" mandate. You'll develop a real POV on how AI changes code review, testing strategy, PR standards, onboarding, and what skills you hire for. You've led teams through this transition before and have specific, battle-tested opinions on what changes and what doesn't. You'll be setting the standard, not delegating it.

Close the gap between engineering output and patient outcomes. We have patient funnel metrics (intake-to-match, match-to-book, book-to-retained) but engineering doesn't yet co-own them tightly enough with Product and Data. You'll establish the operating model where engineering, product, and data science jointly own these metrics — a true triad, not one where engineering is an execution arm.

Evolve team structure as the product evolves. The boundaries between ranking, activation, and onboarding will shift as we move toward intelligent matching. You'll evolve team topology, ownership boundaries, and technical interfaces as the product changes shape — while scaling from ~18 to ~25+ engineers without losing velocity or quality.

• →Set AI and engineering strategy for the core patient experience in partnership with Product and Data Science leadership
• →Lead 3 Engineering Managers; scale the org from ~18 to ~30+ engineers over the next 18 months
• →Co-own patient funnel metrics with your Product and Data counterparts — not just deliver against them
• →Drive delivery of ML-powered matching, reimagined patient onboarding, and patient activation systems
• →Own the AI product roadmap for your pods: where we use ML, where we use LLMs, how we reason about explainability and patient safety
• →Build an engineering culture that uses AI in the workflow and builds AI into the product — as two distinct, equally important practices

• Experience building LLM-based product features (conversational interfaces, intelligent triage, AI-assisted workflows) — this is where patient-facing AI is heading and we want someone who has been there
• Experience rethinking team structure or hiring profiles in response to AI productivity gains — you've thought through what a high-performing team looks like when AI is a meaningful part of how code gets written
• Healthcare experience or other regulated industries where data sensitivity and clinical consequences raise the stakes
• Experience with marketplace dynamics (supply/demand balancing, multi-sided incentive design)

Python (Django/FastAPI) and TypeScript/React on the frontend. Elasticsearch powers search and ranking. PostgreSQL and Redis handle data storage and caching. We use dbt and Snowflake for data pipelines, Temporal for workflow orchestration, and custom ML models for matching. Everything runs on AWS.

For AI development, we use Claude Code and Cursor across the engineering org and are actively evolving our standards for AI-assisted workflows. You'll be setting the direction here, not inheriting a finished playbook.

You won't be writing code daily, but you'll engage deeply enough with these systems to make sound technical and organizational decisions.

• Build AI products with genuine clinical impact — where better matching means better mental healthcare for real people
• Lead the transition to AI-era engineering in practice, not just in principle — with the autonomy to define what that means for your teams
• Co-own patient outcomes, not just engineering output
• Shape an engineering org during a foundational transition, with strong executive support and a clear mission
• Work at a company where the mission isn't marketing copy — patients are actually getting access to therapy they couldn't get before