> Markdown version of [/jobs/ext/3790910-ai-engineer-agentic-systems](https://www.wearedevelopers.com/jobs/ext/3790910-ai-engineer-agentic-systems). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# AI Engineer (Agentic Systems)

- **Company:** Data Intellect
- **Location:** Belfast, UK
- **Experience:** Expert
- **Contract:** Permanent contract
- **Skills:** Java (Programming Language), Application Programming Interfaces (APIs), Artificial Intelligence, Code Review, Databases, Continuous Integration, Cursor, Python (Programming Language), Oracle (Applications), Microsoft Copilot, Software Engineering, TypeScript, Large Language Models, Claude Code, Agentic-AI, KDB+, Machine Learning Operations, Vertica, Claude, Software Version Control
- **Published:** October 17, 2026
- **Apply:** https://www.careerjet.co.uk/jobad/gba07533662ebee3f67d8f04be74b3b0e4

## About the Role

* 5+ years' commercial experience in a Software Engineering role, with a proven track record of designing, building, testing and deploying production-quality software
 * Demonstrable experience building AI agents that use LLMs, cloud hosted or local. We will want to talk in detail about something you have built, why you built it that way and what broke
 * Fluent use of AI-assisted coding tools (Claude Code, Cursor, Copilot or similar) in day to day development
 * Strong software engineering fundamentals: version control, testing, code review, CI/CD and an understanding of what makes code someone else can maintain
 * Strong written and spoken communication. You can write a clear client-facing document and hold your own in a technical conversation with a client
 * Able to work with minimal supervision. You take a problem, ask the right questions early and come back with progress rather than waiting to be unblocked
 * Genuine interest in where agentic AI is going, evidenced by what you read, build or contribute to outside of what your last employer asked of you

Desirable

 * Claude Certified Architect (Foundations or Professional), or an equivalent credential from another vendor. If you have not done one, we will support you through it
 * Background in financial services, market data or another regulated, data-heavy domain
 * Experience with MCP server development
 * Experience with time series or analytical databases (kdb+/KDB-X, ClickHouse, Oracle)
 * Python, Java and TypeScript
 * Experience deploying models on-premise or in a locked-down cloud environment

Who does well here

 * People who are honest about what they do not know.
 * People who put in the work. Consultancy has peaks and clients notice who shows up during them
 * People who can hold a relationship as well as a codebase. Much of our repeat work comes from a client trusting a specific engineer.
 * People who collaborate well and share what they learn, rather than keeping the clever bit to themselves.

## Description

We build AI systems that go into production. Our clients are Tier 1 banks and hedge funds with real data volumes and real regulatory pressure and they hire us because we can put working software in front of them quickly.
This role is for an engineer who has already built agentic AI systems and wants to do more. You will design and build agents that call tools, query client data, reason over it and hand back something a user can act on. Some of that runs against cloud model APIs, some of it runs on local models inside a client's network because the data cannot leave the building. You need to be comfortable with both.
You will work directly with client engineers and stakeholders, as an AI specialist.
What you will be doing

 * Building and shipping LLM-backed agents: tool use, MCP servers, retrieval, orchestration between specialist agents, evaluation and guardrails
 * Working across cloud model APIs and locally hosted models, and making sensible calls about which fits a given client's constraints
 * Using AI-assisted development tools as a core part of how you work, and having a view on where they help and where they get in the way
 * Turning a vague client problem into a scoped proof of concept, then into something maintainable
 * Explaining your design decisions to people who are technical and non-technical people.
 * Building reusable accelerators so the next engagement starts further along than the last one