VP of Engineering (hands-on)

Headways; San Mateo (CA); On-site; USD 180,000 - 280,000; Full time

Benefits offered by this job

Equity stake

Ownership

Job summary

Headways seeks an engineering leader to own the codebase and technical roadmap, writing code daily while guiding the engineering team. You will deploy on-site with customers, gather feedback, and translate it into platform changes that improve velocity.

You will champion security, data governance, and data integration as core product aspects while growing a small, high-trust team that builds AI-native software for enterprise customers.

Qualifications

• Designed and shipped complex enterprise software.
• Led or influenced a small team.
• Experience with on-site deployments and customer presence.

Responsibilities

• Own the codebase and technical roadmap; write code daily.
• Improve product velocity and prioritize scope vs speed.
• Deploy with customers on-site and gather feedback for next changes.
• Ensure security, governance, and data integration are central to the product.
• Build and manage the engineering team as it grows.
• Foster an environment where engineers and builders collaborate safely.

Skills

TypeScript

Full-stack

Kubernetes

Terraform

Data integration

Security and governance

Tools

Postgres

Prisma

SpiceDB

GCP

GitHub Actions

Job description

Headways helps enterprises with their AI transformation. Our Field Deployment Engineers embed inside a customer, capture how the best people actually do the work, and codify it into AI routines that keep running inside the organization, governed, auditable, and measurable. We're deploying at technology and healthcare companies today, with six-figure contracts, a healthy pipeline, and Tim Mayopoulos (former CEO of Fannie Mae) as an advisor.

We're hiring our first engineering leader. If you are someone who is opinionated about how AI-native software gets engineered, wants to own a codebase and is excited to grow a team,

What you'll own

• The codebase and the technical roadmap. You decide what gets built, how, and in what order. You write code most days. This is not a managing-from-the-side role and won't be for a while.
• Product velocity. Our team shipped a production platform by building with AI-first. Your job is to make that faster and more focused with customer feedback. We'd rather cut scope than cut speed.
• Deploying with customers. You'll be on site and in the room when we deploy: watching how a customer's team actually uses what we built, taking the feedback, and turning it into the next platform change. Nobody here builds blind. The roadmap comes from deployments, and you'll have seen the problem before you fix it.
• High-trust deployment. Our customers hand us their Salesforce, Epic, SharePoint, and Teams data, and some of it is PHI. Security, governance, and data integration are the product, not a checklist.
• The engineering team. You'll hire the first engineers and set how they work. Early on that's one or two people; it grows with the customer base.
• "Everyone is a builder." We want support and FDEs building routines and connectors themselves. Engineering's job is to draw the boundaries: what's safe for non-engineers to build, what needs review, and what stays in the core. You'll design that system.

What's already built

This is a real platform, not a prototype. You inherit:

• A TypeScript monorepo (Turborepo, pnpm) running at app.headways.ai: Hono API, React Router 7 product app, DBOS background workflows, an agent runtime with sandbox and egress policies, an ingest service, and connector adapters for the systems our customers run on.
• Postgres with Prisma and SpiceDB for the authorization graph.
• Production on GCP: GKE with per-service Helm charts, Terraform, Secret Manager, Falco, pgcat, Metabase and dbt for analytics. CI/CD through GitHub Actions with Postgres-backed test runs.
• An AI-native development workflow: spec-and-plan docs in the repo, agent-driven implementation, and conventions written for both humans and models.

Some of it is over-built for a team this size, and there's dormant code to retire. Part of the job is deciding what to simplify and what to delete.

Who you are

• You've designed and shipped complex enterprise software, and you've been the person who owned the whole system, not one service of it.
• You like being in front of customers. Not selling, but watching people use the thing, hearing what's wrong, and fixing it. If you'd rather never leave the editor, this will be a frustrating job.
• TypeScript full-stack is your home. You've run Kubernetes in production and can read Terraform well.
• You've done real data integration work: connectors, ETL, permissions, the ugly parts.
• You take security and governance seriously because you've been on the wrong side of a breach, an audit, or a customer security review.
• You've led a small team before, or you've been the obvious person a team followed, and you want to do it properly this time.
• You're fine with uncertainty but not looking for a science project. There are paying customers, a working system, and a pipeline. You want to build, see the growth, and lead.

Healthcare or agent-runtime experience is a plus, not a requirement.

Details

• Compensation: salary, very meaningful equity, and ownership.
• Reports to: Michael Ke Zhang, CEO. You'd work daily with Eric Smejkal, Head of Product, who also ships code.