Forward Deployed Engineer

You are the engineer Bold AI sends into a client's business to make things work. Not a prototype that dies in a demo, not a proof of concept. Production systems that run inside the client's real environment, connected to their real tools, used by their real staff. You spend most of your time wiring systems together, a quarter of it writing code, and a quarter of it in front of the client. If any of those three make you uncomfortable, this is not your role.

• Take a loosely defined client brief and turn it into a scoped, buildable plan within the first week of an engagement.

• Build and ship integrations into live client environments: their data, their access controls, their uptime expectations.

• Deploy to production and stand behind it. Monitoring, error handling, rollback plans, and runbooks are part of the deliverable, not extras.

• Run your own engagement: timeline, risks, client communication.

• Demo working software to clients weekly. Progress is shown, not reported.

• Hand off cleanly at the end: documentation, credentials transferred, a named client-side owner who can run what you built.

• Turn what you learn into reusable templates and playbooks so the next engagement starts faster.

This job has pressure, and you should know that walking in.

• Deadlines are client deadlines. They are tied to launches, events, and budgets. They move rarely and never because the work was harder than expected.

• You will work in unfamiliar territory constantly. A stack you have never touched, an API with no documentation, a legacy system nobody at the client fully understands. Figuring it out is the job.

• Production means production. When something you shipped breaks at a client, you are the first call. You fix it, then you fix the reason it broke.

• Scope will shift mid-engagement. Clients discover what they actually need once they see working software. You adapt without losing the timeline or your composure.

• You represent Bold. On-site and on calls, you are the company. Prepared, direct, and calm, especially when things go wrong.

In exchange: no layers of management, direct access to leadership, real ownership of real outcomes, and work that lands with clients in weeks. You will compress years of experience into each engagement.

Required:

• Production experience. You have shipped systems that real users depended on, and you have been on the hook when they broke. You can talk specifically about an outage or failure you handled and what you changed after.

• Integration fluency. REST APIs, webhooks, OAuth flows, rate limiting, retries, idempotency, queues. You have connected systems that did not want to be connected, and you know where these projects go wrong.

• Solid coding in Python or TypeScript. Not framework trivia. The ability to write a reliable service, script, or automation quickly and leave it maintainable.

• Working infrastructure knowledge. Linux, containers, environment management, secrets handling, basic CI/CD. Enough to deploy and operate what you build without waiting on someone else.

• Client-ready communication. Clear written English, confident on calls, able to deliver bad news early and plainly.

• Composure under pressure. Evidence that you have kept delivering when timelines were tight, requirements moved, or something was on fire.

Strong assets:

• AI/LLM integration experience: model APIs, tool/function calling, MCP, RAG fundamentals.

• Prior agency, consulting, or forward-deployed experience.

Experience inside enterprise client environments: SSO, VPNs, security reviews, procurement-grade documentation.

50% Integrations. This is the core of the job. Connecting client systems to each other and to AI: CRMs, ticketing systems, Slack, email, databases, internal tools, third-party APIs. Reading someone else's API docs, handling their broken auth flow, working around their rate limits, and making data move reliably between systems that were never designed to talk to each other. Most of what we deliver is not new software. It is existing systems, finally connected, with AI in the loop.

25% Writing Code. The glue, the middleware, the automations, and the occasional custom interface. Python or TypeScript, small services, webhooks, scheduled jobs, data transforms. Clean enough that another engineer can maintain it, fast enough that the client sees progress every week.

25% Client Facing. Scoping calls, weekly check-ins, demos, and the hard conversations when scope shifts or something breaks. You explain technical decisions to non-technical people without jargon and without hiding behind it. The client should know your name and trust it.

• A practical exercise built around a realistic integration problem, not a whiteboard puzzle.

• A deep walkthrough of one production system you shipped: decisions, failures, fixes.

• A scenario conversation: a client call where scope is shifting and the timeline is not. We want to see how you think and how you communicate, not whether you have a script.