The posting
About Indicium AI
Indicium AI was Anthropic's first European launch partner, and sits among a handful of organisations globally trusted by Anthropic to deploy Claude at enterprise scale. We are also global AWS, Microsoft and Databricks partners and Databricks Ventures has invested in our business and holds a seat on our board. Our engineers get access to the latest AI tools and models, partner technical teams, and early capability previews that the wider market does not.
We are deploying production AI systems into the world's most complex organisations across Financial Services, Energy, and Healthcare. We exist to make AI a genuine competitive advantage for our clients, not in a presentation but in production.
We invest seriously in our people. Every FDE at Indicium AI has access to Anthropic training, and a flexible learning budget to go with it. All of our staff are equipped with Claude licenses for Chat, Cowork and Claude Code. We are building a team of engineers who work at the frontier, and we back that with the investment to match.
The Role
This is a role for someone who wants to build. You will be embedded in a client team, shipping production agentic AI systems that real users depend on, including multi-agent orchestration, RAG pipelines, MCP server integrations, and the evaluation and observability layer that makes non-deterministic systems trustworthy in regulated environments.
As a Senior FDE, you are the person the team brings hard technical problems to. You will take ambiguous problems, scope and prioritise them, then work with your squad to deliver high quality agentic AI systems into production at pace. You do not need a full spec to start. You need access, a clear outcome, and the judgement to know when to ask.
You will be in front of clients regularly, explaining what you built to their engineers and communicating the business value to non-technical stakeholders. You will build the kind of credibility where your technical judgement gets taken at face value.
What You Will Do
- Build and ship production-quality agentic AI systems end to end; agent orchestration, tool-use patterns, RAG and retrieval architectures, and the reliability layer around them
- Design and implement evaluation frameworks and observability for non-deterministic systems
- Raise build quality deliberately through testing, output validation, and automating manual checks
- Earn the trust of client-side technical leaders as someone who delivers complex things and tells the truth about them
- Stay current as the field moves, and bring what you learn back to the team
- Explain trade-offs clearly to non-engineers, and manage expectations honestly
- Improve how the whole team works, not only your own output
- Contribute reusable assets, reference implementations and knowledge back from client work. The things that make the next engagement faster
We'd Like To Speak To You If...
Read this as "some of these", not "all of these." Almost nobody has the full list. If several sound like you, and the rest sound like things you would enjoy getting good at, we encourage you to apply.
- You have built and deployed agentic AI systems into production, not just POCs or personal projects
- You have worked with multi-step LLM workflows including tool use, multi-agent orchestration, using frameworks such as the Claude Agent SDK, LangGraph, Strands or CrewAI, or your own implementation where a framework did not fit
- You have implemented RAG or retrieval architectures: embedding strategies, chunking, hybrid search, reranking, vector or graph stores
- You write production-grade code comfortably, and review other people's code to a high standard. Python is our common language, though strong fundamentals in another modern language is interesting to us too
- You have evaluated LLM systems and have opinions about how it should be done in enterprise settings
- You are comfortable with at least one major cloud (AWS, Azure, GCP) and have shipped into production environments with real enterprise constraints
- You have worked with clients or stakeholders outside your own team, and can hold a technical conversation with people who are not engineers
- You take ownership in ambiguity. You make a call, communicate it, and deliver, rather than waiting for the brief to arrive
Also Interesting But Not Required
- MCP server development and tool ecosystem integration
- Experience in Financial Services, Energy, Healthcare, or another regulated environment
- Model observability tooling such as LangSmith or Weights & Biases, and cost and latency optimisation at scale
- Writing, speaking, or community contribution in AI engineering
- Experience working alongside partner ecosystems such as Anthropic, Databricks, or AWS
What we are not asking for: a specific number of years, a particular degree, consulting experience, or hands-on time with every tool listed above. We are asking for evidence that you build things that work, and that you learn quickly.
Why Indicium AI
Access to leading AI Labs: Anthropic’s first European launch partner. One of the first four OpenAI UK partners. Databricks Ventures invested and on our board. Our engineers work with capabilities before the wider market does and work jointly with teams from leading AI Labs on customer delivery.
We are builders: We embed, we architect, we build - and we stay until it works. Our engineers make real decisions on live systems, for clients in industries where getting it wrong has consequences. There are no layers between you and the work.
Problems worth solving: No two engagements are the same. The variety and complexity of enterprise AI problems at Indicium AI is something a product company simply cannot offer. If you want engagements that test every dimension of what you can do, this is where they are.
Investment in staying at the frontier: The best engineers we know are relentlessly curious - and we build an environment that supports learning and growth. Every engineer has access to leading AI training and tools, career development pathways, and a flexible L&D budget. We think that’s how it should be.
