Forward Deployed Engineer

We are building the FDE function at Diligent from the ground up with ambitions to grow this team to 20–25 people. The first three hires will shape how we embed AI agents into some of the world’s most complex governance, risk and compliance environments. This is not a support or consultancy role.

It is a builder role, for someone who is equally comfortable reading a failing agent trace, running a discovery workshop with a bank’s internal audit team, and translating what they find into a production-grade agentic solution. You will be building AI agents for GRC professionals , not assistants that surface suggestions, but agents that own complex, multi-step workflows end to end . Agents that customers can hand a task to and trust it will come back done.

Closing the gap between a promising prototype and something a company Board and ELT depends on is a completely different discipline to building the prototype. That is what this role is about.

What You’Ll Do

Embed directly with major enterprise customers (global banks, regulated corporates) across EU and US , sitting with internal audit teams, risk functions, compliance and governance professionals to understand their real workflows, not to demo what the agents already do. Run agent-focused discovery workshops, rapidly prototype agentic solutions, and test them with practitioners; distinguishing between workflows that need an agent and those that need a button. Source, integrate, and move data between enterprise systems as part of live customer implementations — understanding the real data landscape customers operate in and building reliable pipelines to support it .

Take agents from prototype through to production-grade reliability: building evaluation infrastructure, golden datasets, guardrails, and observability so a compliance team can trust the output. Master the hard failure modes of agentic AI — silent regressions on model updates, context window degradation, prompt instability, non-deterministic outputs — and build the infrastructure that prevents them. Synthesise learning across multiple enterprise accounts to identify which agent behaviours should be generalised into the platform, feeding field insights back to product and engineering.

Requirements

, and agent tool specifications for internal teams. These are the essentials you’ll need to get an interview Hands-on experience shipping at least one SaaS production agent from prototype to evaluation to live deployment to regression — and the scars to prove it. Proven implementation experience: you have worked on enterprise deployments where you have sourced data from multiple systems, built integrations, and onboarded complex customers onto technical platforms.

This is a hard requirement. Deep understanding of the Agent Development Life Cycle: evaluation frameworks, guardrails, observability, prompt versioning, and golden datasets. Strong software engineering fundamentals: APIs, data pipelines, backend services, agent tool-calling frameworks (MCP or equivalent).

TypeScript and/or Python. Experience with multi-agent orchestration patterns: orchestrator/sub-agent architectures, agent-to-agent coordination, shared context models. Genuine comfort operating across both technical and business environments — as effective with a company Board and ELT as you are debugging a failing agent trace.

Curiosity about GRC, audit, and compliance domains. You don’t need a compliance background, but you need the drive to learn it fast. Flexibility to travel regularly across EU and US customer sites.

and technical delivery are often a natural fit for this work. Familiarity with GRC, audit, or compliance software platforms.