Forward Deployed Engineer

Why this role? We build business systems for executives at mid-market companies. The work runs across industries and across the executive team, reaching the CEO, COO, CFO, and CRO, through three practices: fractional CFO leadership, AI enablement, and revenue engineering. Forward deployed means what it says. You do not build in a sandbox and hand over a repository. You build in the client's environment, against their identity provider, their network rules, their data, and their approval chain, with their people watching. The analyst seat writes the specification. This seat turns it into something running, then stays until the client's own team can run it without you. Consulting background is a strong signal here. Time at Accenture, Deloitte, McKinsey, Bain, BCG, Slalom, or a comparable delivery organization means you already know how to hold a room, drive a workstream to a date, and write a document that serves an executive sponsor and an engineer at the same time. We weigh that as heavily as the code. What you would own Seven areas. The first engagement is the reference rebuild. From there the pattern extends into new client work and you keep it running. Deployment inside client environments Stand up integrations, data pipelines, and AI workflows in systems you did not design and cannot rebuild. Work against real constraints: SSO and role models, network and residency rules, change windows, an ERP nobody has touched in four years. Ship in increments the client can see. Solution engineering from a signed specification Take a Functional Requirement Document and produce working software. Draw the line between platform configuration and custom code, and defend where you drew it. Write the integration, the transformation, and the service that holds it together. Agent design and orchestration Build agent workflows with retrieval, tool calls, and evaluation attached. Place the human approval gate deliberately and log what crosses it. Set cost and latency budgets before the build, then measure against them. Data readiness and migration Profile what is actually in the source system, not what the client believes is in it. Set mapping, migration, and quality rules. Nothing moves until the reconciliation ties. Project management and delivery cadence Own the plan, the dates, and the risk register on your workstream. Run a weekly status that a sponsor can read in ninety seconds. Raise a slip the week you see it, not the week it lands. Executive presentation and client communication Demonstrate working software to a CFO or COO and translate the technical tradeoff into the business decision they have to make. Run the working session that ends in a signed-off answer. Write the one-page recap the same day. Enablement and handoff Build runbooks, architecture documentation, and standard operating procedures. Train the client's team on the system they now own. Carry hypercare through the first close, the first quarter-end, or the first campaign cycle, whichever proves it. Who this suits Engineers who want the client in the room. If you would rather take a ticket, close it, and never meet the person who filed it, this is the wrong seat. You will demonstrate unfinished work to a CFO and explain what it does not do yet. You will make an architecture call on a Tuesday with incomplete information and write down why. And when a deployment goes sideways, you say so first, with the fix already drafted. Engagement terms Work starts on a defined scope with a written statement of work, usually one client deployment. Good work earns the next engagement. As the practice grows we expect this seat to become full-time. No date is attached to it, and we will state plainly where things stand at each renewal. What we expect Five years building and shipping production software, including at least two deployments run inside a client or customer environment through go-live. Engineering Python and TypeScript or JavaScript, REST and webhook integration patterns, source control, CI, deployment to a live environment. 5+ YEARS AI systems LLM application work in production. Retrieval, tool use, agent orchestration, evaluation, guardrails, cost and latency control. 2+ YEARS Data SQL, warehouse work on Snowflake, BigQuery, or Postgres, migration and mapping, quality profiling, reconciliation. 3+ YEARS Platforms API and configuration depth on Salesforce, HubSpot, NetSuite, SAP, or comparable enterprise systems. 3+ YEARS Project management Owning a workstream to a date. Scope control, dependency tracking, status reporting, early escalation. 3+ YEARS Communication Writing for a sponsor and an engineer in the same document. Presenting to an executive audience and defending a recommendation live. 3+ YEARS Security and controls Authentication, secrets handling, least-privilege access, audit logging, PII treatment inside client systems. WORKING DEPTH Useful, not required Time at a consulting or systems integration firm. Delivery discipline under a client clock carries over directly, and so does the habit of writing things down. Forward deployed engineering, solutions engineering, or professional services at a software company. Control mapping against NIST AI RMF, ISO/IEC 42001, or SOC 2, and comfort talking to an assessor. Finance or revenue systems depth: quote-to-cash, revenue recognition under ASC 606, the monthly close, CRM and CPQ architecture. How to apply: Send the following to gena@f3insights.com with the subject line "Forward Deployed Engineer". One thing you built and deployed inside somebody else's environment. A repository, an architecture diagram, or a written account. Redact client names and figures. A resume or profile link. No cover letter. A short note on one deployment that went wrong: what broke, what you changed, and what you now do differently. Selection process: 01 Screening call, 30 minutes. Scope of the first engagement, your availability, and rate. 02 Working session, 60 minutes. Walk us through a deployment you led, then a live architecture and scoping conversation against a sample client situation. Expect to present, not just answer. No take-home project. 03 References and paperwork. Two references, then a statement of work for the first scope. We aim to give a decision within a week of the working session. Questions about the role before applying go to the same address.