About This Role
As a Forward Deployed Engineer, you'll embed directly with client teams to rapidly ship AI systems that drive measurable business outcomes. You're not building in isolation—you're working alongside client stakeholders, understanding their problems firsthand, and delivering production-ready solutions in weeks. This role combines the speed of prototyping with the rigor of production engineering, requiring someone who can navigate ambiguity, ship v0.1 quickly, iterate based on user feedback, and leave behind durable, well-instrumented systems that clients can maintain and extend.
Mission
Embed with client teams to ship AI systems that move real metrics in weeks, not quarters.
Outcomes You'll Own
- Translate ambiguous business goals into production systems (agents, workflows, integrations, data pipelines) with measurable ROI (time saved, error rate, revenue lift)
- Stand up secure services (APIs, workers, event pipelines), instrumentation, evaluation harnesses, and observability
- Leave clients with durable capability: docs, handover, and re-usable components
What You'll Do
- Rapid discovery → prototype → iterate → productionize
- Build with Python/JS (FastAPI/Flask/Node), cloud (AWS/GCP/Azure), vector/RAG patterns, agents/orchestration, and front-end surfaces (React)
- Integrate third-party APIs/CRMs/ERPs; design data contracts and quality checks
- Implement evaluation: unit tests for prompts/tools, regression suites for agents, offline/online A/B
- Handle security, privacy, and access patterns pragmatically
You'll Thrive If You
- Love shipping v0.1 and v1.0; talk to users daily; delete code happily
- Can reason across product + data + infra; comfortable with incomplete info
First 90 Days
- 30 days: Deliver a working vertical slice with tracked ROI
- 60 days: Land a second use-case; add evaluation + observability
- 90 days: Package common components into a reusable kit; mentor teammates
Example Projects
- Claims triage agent across email + EDI + CRM with human-in-the-loop
- RAG over private manuals with continuous ingestion and feedback training
- Sales ops automations tying LLM agents to CPQ + contract generation
Responsibilities
- Rapid discovery → prototype → iterate → productionize
- Build with Python/JS (FastAPI/Flask/Node), cloud (AWS/GCP/Azure), vector/RAG patterns, agents/orchestration, and front-end surfaces (React)
- Integrate third-party APIs/CRMs/ERPs; design data contracts and quality checks
- Implement evaluation: unit tests for prompts/tools, regression suites for agents, offline/online A/B
- Handle security, privacy, and access patterns pragmatically
- Stand up secure services (APIs, workers, event pipelines), instrumentation, evaluation harnesses, and observability
- Leave clients with durable capability: docs, handover, and re-usable components
Requirements
- Strong Python and/or JavaScript; comfort with cloud, data stores, APIs, and CI/CD
- Experience taking scrappy prototypes to reliable services
- Can reason across product + data + infra; comfortable with incomplete info
- Love shipping v0.1 and v1.0; talk to users daily; delete code happily
- Bonus: retrieval tuning, embeddings, structured output, function/tool calling, queue systems, message buses, GPUs