You lead delivery on a single client engagement, start to finish. Discovery, scoping, system design, build, and production rollout. You write the code, work directly with the client, and stay accountable for what ships.
This role takes three things in one person: the judgment to understand the problem, the ability to talk it through with the client, and the skill to build what solves it. Most people bring two. We need all three.
A large part of the job is translation. You sit in operator conversations, hear what the business actually needs, and turn it into scope, deliverables, and acceptance criteria the client agrees to and the team can build against. The client trusts you to hold that line.
Success is measured by production adoption and workflow impact, not hours logged or demos given.
What you'll own

• The full delivery lifecycle for the client engagement, from first prototype to stable production.
• Full-stack build across Python and JavaScript/TypeScript, plus the data and integration work that connects to client systems (APIs, third-party platforms, accounting and telemetry sources).
• AI features built on top of Client's core work. You integrate LLMs into the product and ship them; our core team owns the deeper agent design and evals.
• Direct client relationships. You translate operator needs into scope, set expectations, and drive adoption of what you build.
• Turning client business logic into durable software that holds up after you hand it off.
• Sequencing work, surfacing risk early, and making the scope, speed, and quality trade-offs that protect the delivery date.

You might fit if you

• Turn ambiguous client conversations into clear scope, deliverables, and acceptance criteria, and hold that line as work moves.
• Communicate directly with operators and non-technical stakeholders, and earn trust as the person they call.
• Have 5+ years building and shipping software, including client-facing delivery.
• Work well in ambiguity. You scope, decide, and move without waiting to be unblocked.
• Close loops. Ownership over activity.

Technical skills
Must have

• Production full-stack: Python on the backend, TypeScript and React on the front. Ships in both, not familiar with one.
• Comfortable building against LLM APIs and wiring them into an application. You don't need to architect agent systems; Client supplies that depth.
• API and third-party system integration, including pulling and cleaning messy source data.
• SQL and data modeling.
• Cloud deployment and CI/CD. Owns the path to production.

Plus

• Deeper agent design and eval experience. A bonus, not a requirement, since our core team owns that work.
• Containers and infrastructure-as-code, useful on the platform side of the work.
• Monitoring and observability.
• Direct experience with accounting, field operations, or telemetry data.