Forward Deployed Engineer

About the role

We practice Ideation Engineering. We embed with enterprise clients to turn their toughest problems into AI systems that actually ship. As a Forward Deployed Engineer you are the person on the ground, part engineer, part consultant, turning the real constraints of their environment into working software.

You work the way we build: agents at scale, humans accountable. You pair with coding agents on every story, own the context layer they run on, own every merge. You set the approach, review the diff, hold to the eval thresholds a release has to clear. The agents carry the volume; you are accountable for what ships. This is a high-ownership, client-facing role that suits someone who thrives on open-ended problems and measures their work by what reaches production.

What you'll do

• Embed with enterprise client teams, remote or on-site, as the credible technical voice with their engineers, product leaders, and executives.
• Own every merge. Pair with coding agents on each story, set the approach, review the diff, stay accountable for what ships.
• Translate the PO's product brief into an engineering design (a versioned EDD.md). Scope what agents may do and where they may act, then stand up the infrastructure it runs on.
• Build and version the context layer: the rules, data, tools agents use to reach a client's systems (MCP servers, internal APIs, integrations across GitHub, Jira, Linear, Slack, Sentry).
• Pick the right model for each job from a fleet of frontier and open-weight options, then run it where it belongs: a vendor API, an open-weight model on vLLM inside the client's VPC, or on-prem for regulated work.
• Own the eval harness that proves behavior. Write the evals, calibrate them against the thresholds the PO sets.
• Prototype AI-powered solutions quickly (agentic tooling, automation), iterating with stakeholders in tight feedback loops.
• Harden prototypes into reliable, observable systems that the client's own engineers can run after you leave.

What you'll bring

• 5+ years building and shipping production software. You've owned systems in production, not just written features.
• Strong engineering ability across our stack: Python, C# / .NET, TypeScript (Java, Go, SQL a plus). You are comfortable on frameworks like ASP.NET Core, FastAPI, Node.js, React / Next.js.
• A working knowledge of the model landscape, and when to use each: frontier families (Claude, GPT, Gemini, Grok) alongside open-weight models (Llama, Qwen, DeepSeek, Mistral, gpt-oss). Comfortable running them via vendor APIs, vLLM, or Ollama.
• Hands-on experience with agentic coding tools (Claude Code, Codex, Cursor, OpenCode), plus the context and eval tooling (MCP servers, eval harnesses) that makes agents reliable.
• Practical experience with LLM application patterns: retrieval, agentic workflows, evaluation, prompt and context engineering.
• Cloud fluency (AWS, Azure, or GCP) with a solid grasp of CI/CD, containers (Docker, Kubernetes), infrastructure as code (Terraform), and databases (Postgres, Supabase, Neon).
• Clear communication. You can shape a loose brief into a clear scope with a stakeholder, then explain a trade-off to a non-technical executive.
• Comfort with shifting priorities and a fast client-facing pace. You default to shipping, own the outcome, travel occasionally for client embeds.

Bonus points

• Consulting, agency, forward-deployed, or solutions-engineering experience.
• Experience serving open-weight models in a VPC (vLLM) or on-prem, or building MCP servers and agent context layers.
• A track record of modernizing legacy enterprise platforms.
• Exposure to financial services or healthcare.