Forward Deployed Engineer (AI-Native)

Responsibilities

Embedded build (50%)

• Work directly with the function's data, documents, systems and APIs; learn the workflow by sitting with the people who run it.
• Ship prototypes and MVPs in days using AI-assisted development, and reship the same day when feedback lands.
• Build production agents: multi-step workflows, RAG over internal content, document extraction, tool use, and integrations with ERP, CRM, CMS and contract systems.

Run and improve in production (25%)

• Own evaluation suites, guardrails, observability, latency and token cost for every agent you ship.
• Track adoption and business outcomes (hours saved, turnaround time, error rate) and iterate on them.
• Handle incidents and model or prompt regressions; keep a clear change log.

Translate and enable (25%)

• Be the technical presence with business leaders and users; surface risks early and turn business language into technical scope.
• Run demos and training so teams get real value from the agents built for them.
• Document systems, decisions and gotchas, and add reusable patterns to the team playbook.

Skills Required

• 3+ years of software engineering, including 1+ year building agentic systems that real users rely on. Prototype-only experience does not qualify.
• Strong Python; working full-stack ability in TypeScript / Next.js, PostgreSQL and REST APIs.
• Hands-on with agents: tool use, orchestration, RAG, embeddings, vector databases, evaluation and guardrails.
• Experience with Gemini Enterprise / Vertex AI on Google Cloud (Cloud Run, Cloud Functions), or AWS Bedrock with the ability to ramp up on Google Cloud quickly.
• Uses AI coding tools (Claude Code, Cursor or similar) as a daily default.
• Comfortable with ambiguity, undocumented systems, and rough demos with live feedback.
• Explains AI clearly to non-technical colleagues and writes short, clear documentation.