Headway

Eight weeks ago we built a system of AI agents at Headway to clear out our engineering backlog and proactively find and fix issues. Since then it's shipped over 250 pull requests. None of the code was written by a human (although it was all reviewed by one). Our backlog had over a thousand items, some older than three years, and every one of them was a small papercut but never mattered enough to beat the next feature off the roadmap. Slow pages. A broken link nobody clicked. Tests that failed for reasons unrelated to the code. Essentially the broccoli of engineering. So Will Anderson and I built something to eat the broccoli for us. Agents autonomously find bugs, write the fix, open a PR, then babysit it through CI and code review until it's ready. What's left for the engineer is the part that actually needs a person: looking at the diff and deciding whether it ships. I went in assuming the hard problem was getting an AI to write decent code, but it wasn't. The hard problem was orchestrating dozens of agents to work in parallel without colliding, without quietly hallucinating, and without drowning reviewers in PRs nobody asked for. Turns out building autonomous swarms is guided by the same principles that make human teams work: keep the scope defined, make ownership obvious, give everyone a structured place to coordinate. My colleague Will and I wrote up how it's built, where it broke, and what running it in production for two months actually taught us. Read the article here: https://lnkd.in/e4ZuU6eR If you’re interested in building technical healthcare infrastructure and creating the future of mental healthcare in America, we’re hiring!

Arav Misra and I spent the last few months building autonomous coding agents as a side project at Headway. The big takeaway? They don't just need prompts, they need strict structure. Our first version pulled from free-text Slack messages and failed hard. Agents duplicated work and misread requests. Everything changed when we moved them into Linear. Using structured tickets with clear fields and ownership labels gave them a reliable "brain." Now, they are completely autonomous. On a set cadence, they read Linear, pick up tasks, and write results back. They even babysit their own PRs, fixing CI errors and responding to reviews. They’ve already merged over 250 PRs, and that number is rising quickly. Zero human-written code, but 100% human-reviewed. This only happened because the Eddy team built the foundation, and Headway and my manager, Robert Chen, gave me the rare space to chase a side project. It is a big reason why I love building here. We wrote this up in a Medium article here: https://lnkd.in/erB8S-PQ If work like this interests you, don't hesitate to reach out!

At Headway, we built an AI system that debugs our dbt failures in production. It started as a hackathon project (and a pun I refuse to apologize for) and ended up freeing our scientists & engineers to build the future of Headway, rather than just troubleshooting the past. The key insight wasn't making the LLM smarter. It was making every failure recoverable, and keeping humans in the loop for the decision, not the investigation. Before this, a 5 AM production failure meant 45 minutes of an on-call engineer's morning: digging through logs, tracing the error, filing a ticket, opening a fix. Now that same incident takes 5 minutes. The system does the investigation, while the engineer reads what happened and decides what to do. At Headway, reliable data infrastructure isn't just an engineering problem, it's what lets our team focus on expanding access to mental healthcare instead of chasing down pipeline bugs at dawn. If your team's debugging playbook fits in a Markdown file, it can probably run itself. Read all about it here: https://lnkd.in/ep7ntQe4

Every mental health provider who joins Headway gets credentialed before they see a single patient. At our scale (hundreds of providers a day) that process needs to be airtight. In the early days of building the company, it ran on spreadsheets. That's how you get a startup off the ground. But credentialing is genuinely complex: external APIs that take days to respond, state-by-state regulatory checks that can run in parallel or depend on each other, and decision points where a human needs to weigh in before anything moves forward. Eventually, the volume and the complexity made it clear we had to rebuild. So last year we did, on Temporal. The system now runs across 12+ teams at Headway and has processed tens of thousands of providers, faster and more accurately than before. Temporal made this surprisingly simple. I wrote about the whole journey: what we tried, what we rejected, and what it looks like when you keep humans in the loop exactly where they need to be. Read it here: https://lnkd.in/gHQMwa8S

We’re honored to be nominated for The Webby Awards, recognizing work that matters. This also puts our work in the running for the People's Voice Award. Check out our work below and give us a vote by tomorrow, April 16. Purpose in practice - Porsche Terry  Health & Wellness (Nominee) ➡️Vote: https://wbby.co/57546N Purpose in practice - Courtney Mazzola  Category: Belonging & Inclusion (Honoree) View: https://wbby.co/57544H Mental health providers are at the heart of Headway. Through our Purpose in Practice campaign, we practiced what we preached. This isn’t about us. After all, it never was. These are their stories; their best practices. We’re just here to empower them. Porsche’s is just one of three captivating therapist stories. You can get a closer look at the entire campaign here: https://lnkd.in/eVAB-sYs

I want to tell you about Eddy (Headway's AI agent/stack): Last month I merged 60 PRs. Not because my calendar suddenly cleared, but because the way I work right now looks nothing like it did a few months ago. AI as a "copilot" is already outdated. When it’s actually wired into your systems (and built with real safety/constraints), it stops being a tool you use and starts being something you work with. You try more things. You ship more. You stop overthinking whether something is “worth the time” and just do it. A small team at Headway decided not to wait around for that to exist. So we built our AI stack. What’s been most interesting isn’t the speed, it’s the behavior change. Engineers aren’t just moving faster; they’re attempting things they would have previously filtered out as “too heavy.” Non-engineers are building workflows that used to require dedicated teams. The line between idea and execution is getting very thin, with that people attempt bolder things. I knew we'd built something special when our internal beta grew from 10 people to 600 in a month. Hard to really explain in just one post, so I wrote more on our blog about what we built, the constraints (especially in health care), and what’s coming next. Read it here: https://lnkd.in/e6iVTtvZ If you’re excited by building systems like this, we’re hiring. #AI #Engineering #Healthcare cc Headway Adam R. Kuhn Francisco Alvarez Matthew Tse Mike Perrone Amish Gupta Hagen Green

Mental health care is deeply human; trust, empathy, and clinical judgment will always be at the center. But providers spend too much time on admin, and patients struggle to find the right care at the right time. When applied thoughtfully, AI can make the experience more human by giving people more space to focus on care itself. That’s why we’re excited to welcome the team behind Tezi, an AI-native company whose builders share our belief that the hardest problems in healthcare require humans and AI working together. Read more in Fierce Healthcare: https://lnkd.in/ey7SacSV Thanks to Cailey Gleeson for covering the news. 

Six months ago, I started spending most of my time on AI to understand what it can meaningfully change about mental health care for providers and their patients. I’ve been consuming millions of tokens a week, stress-testing what's real, and working with our team to figure out where we can make the most impact. Headway has been innovating behind the scenes, and today we're accelerating our investment by acquiring the team behind Tezi, an AI-native company. Mental health is one of the most complex corners of healthcare. Providers spend too much time dealing with insurance, billing, and admin when they could be focusing on patients. Patients, meanwhile, struggle to find the right clinician, understand their coverage, or get an appointment when they need one. The connective tissue between patients, providers, and payers is fragmented in ways that make care harder to access and deliver. We believe humans plus AI will be better than either alone: clinicians supported by intelligent systems, and patients guided to the right care faster. We know we've only scratched the surface of what AI can do for Headway and the people it supports, and are investing heavily to learn the potential. The Tezi team are exceptional builders across engineering, product, and design who share our conviction that the hardest problems require humans and AI working together. They built systems where human judgment and AI collaborate inside complex, real-world workflows, removing the tedious stuff so professionals can focus on the work that actually requires their expertise. That approach maps directly to what we're building at Headway. Raghavendra Prabhu, who founded and led Tezi, is joining as VP of Engineering. He's built and scaled systems at Covariant, Thumbtack, Pinterest, Twitter, Google, and Microsoft. I was blown away by what he and the team built, and I'm thrilled they're joining us to do this work for our 70,000+ providers and the millions of patients they serve. If you want to build at the intersection of AI and health care with a team that's moving fast and thinking big, we're hiring: https://headway.co/careers

Mental health care is one of the defining challenges of our time. Millions of people are trying to find care, understand their insurance, and navigate a system that’s often confusing and fragmented. Solving that isn’t just a clinical challenge, it’s a technical one too. At Headway, our engineering and product teams are working on the infrastructure behind a better mental health system: helping patients find the right provider, making insurance more predictable, and reducing administrative burden for clinicians. The pace and impact of that work has never been greater, so we decided to start writing about it. Today we’re launching the Headway Tech Blog, offering a behind-the-scenes look at what it’s like to build a new mental healthcare system that everyone can access. Our first post, authored by Adam R. Kuhn and Logan Dillard, dives into how we used machine learning to improve healthcare billing accuracy, moving from a massive rules engine to a production ML system that reduced billing interpretation errors by 40% across the platform. Check out the team's first entry here: https://lnkd.in/eX-TBey7

Unlocking access to care is only the beginning. So many questions follow: ➤ Does the patient come back? ➤ Does a therapeutic alliance form? ➤ Do outcomes improve? At Behavioral Health Business' VALUE conference, Headway’s Jennifer Christian-Herman joined leaders from Aetna, a CVS Health Company and Octave to discuss how the industry can move beyond measuring network adequacy toward measuring what actually matters: engagement, clinical progress, and long-term impact. With millions of appointments across our provider network, we believe scale can be a quality engine — helping the industry shift from measuring activity to measuring outcomes. Thank you to Laura Lovett for leading such an important conversation!