Principal Engineer, Client Delivery

Job Description

Our clients bring us messy business problems and expect working software in weeks. You're the technical lead who gets them there.

\n

\n

You'll work alongside our strategists and designers in discovery, then turn what the team learns into an architecture, a plan, and a product. You build with a small team of engineers and AI coding agents, and you know how to get far more out of those tools than most teams do.

\n

\n

Most people who fit have held a Director, Senior Director, VP, or CTO title. They know a lot about a lot, they still build, and they'd rather ship than manage a process.

\n

\n

Vertice Labs builds custom software and AI-native operations for founders, operators, and established companies. Small senior US-based teams of strategists, designers, and engineers. No offshore, no subcontracting. The people on the call are the people doing the work.

\n

\n

What you'll actually do

\n

\n
• Lead the technical side of client engagements, from discovery through launch
\n
• Work with our strategists and designers to find the problem behind the one the client named, and say what's feasible, what it costs, and what to build first
\n
• Make the architecture calls and write real code, with AI agents doing much of the typing
\n
• Lead the engineers on your engagement and set the bar for how they work with AI
\n
• Own the technical conversations with the client, including the hard ones about scope, risk, and tradeoffs
\n

\n

\n

What we require, and how we'll check it

\n

Engineering judgment and people judgment in the same person. You can explain a tradeoff to a CFO in two sentences and to a staff engineer in twenty minutes. You notice when the person nodding on the call doesn't agree. Clients, designers, and engineers all ask to work with you again. Come with examples, because we'll ask for them.

\n

\n

Delivery ownership. You've been the accountable technical lead on a project from first conversation to production, more than once, as part of a cross-functional team. We'll pick one and go five layers deep: the architecture, what you gave up, what broke, and what still bugs you.

\n

\n

Architecture across the whole system. You've designed systems that span frontend, backend, data, and cloud infrastructure, and you can defend the tradeoffs.

\n

\n

Real leverage from AI tools. You use coding agents every day and know where they help and where they need checking. In the interview you'll screen-share your real setup and take a change end to end. We'll watch how you check what the agents produce.

\n

\n

12+ years building software , with the recent ones spent as the technical backbone and the person in the room.

\n

\n

Nice to have

\n

\n
• Consulting, agency, or other client-services background
\n
• Production LLM, retrieval, or agent systems with real users
\n
• AWS/Azure/GCP and Terraform
\n

\n

\n

What this role is not

\n

\n
• Not a hands-off management seat. You lead the technical work and you still build.
\n
• Not a solo seat. You work with strategists, designers, and engineers, and you make them better.
\n
• Not a specialist seat. You own the technical outcome across the whole stack.
\n
• Not a big-company role. Small teams, real ownership, no process for its own sake.
\n

\n

\n

Compensation and location

\n

$225,000 to $275,000 base, depending on experience. US-based, fully remote.

\n

\n

How to apply

\n

Apply with your resume. If you clear the screening questions, we'll send you a short form with these four questions. Be specific. Vague answers are the fastest way out of the pipeline.

\n

\n
1. Tell us about a project you led where the problem the client brought you wasn't the real problem. How did you find out, and what did you build instead?
\n
2. What's the hardest conversation you've had with a client or executive about scope, a deadline, or a mistake? What did you say, and what happened to the relationship afterward?
\n
3. Pick one system you architected that's in production today. What still bugs you about it?
\n
4. What does your AI-assisted dev setup look like right now, and what did you change about it in the last month?
\n