About Scrumconnect Consulting
Scrumconnect Consulting is a multi-award-winning digital consultancy, recognised for
delivering impactful and innovative technology solutions across UK government
departments. Our work has positively influenced the lives of over 40 million UK citizens.
Overview
We’re looking for an AI Engineer to build the application-level AI capability inside a
large-scale AI adoption programme for a central government department. Where the
Machine Learning Engineer trains models and the MLOps Engineer runs the platform, you
sit closest to the end product: turning LLMs, retrieval systems, and agentic workflows into
working, usable features that integrate with the department’s existing services.
You’ll work inside a collaborative “Rainbow Team” alongside civil servants, implementing the
guardrails and standards set by the Lead AI Engineer and Lead Technical Architect.
Key Responsibilities
Build & Integrate Implement RAG pipelines, prompt templates, and agentic workflows, and
wire them into application logic that connects with the department’s legacy and multi-cloud
systems.
Prompt Engineering & Evaluation Design, test, and iterate on prompts and retrieval
strategies; run structured evaluations to track answer quality, retrieval accuracy, and
hallucination rate over time.
Responsible AI in Practice Implement the specific guardrails and human-in-the-loop
checkpoints defined by the leads — this is an execution role, not a policy-setting one.
Documentation & Handover Contribute to “Docs-as-Code” artifacts supporting the OKUA
knowledge-transfer framework, so the features you build can be understood and maintained
by internal government engineers.
Required Skills & Experience
- Practical experience building applications using LLMs or other foundation models, via API or self-hosted
- Experience with RAG architectures, vector databases, and embedding-based retrieval
- Solid general-purpose programming (commonly Python) and API design/integration skills
- Prompt engineering with a systematic, evaluation-driven approach rather than trial and error
- Familiarity with agentic/orchestration frameworks and multi-step AI workflows
- Comfortable working within secure, regulated environments and handling sensitive data appropriately
Desirable Experience
- Experience delivering digital services within UK central government or the wider public sector
- Familiarity with GOV.UK Service Standard expectations
- Exposure to open-source LLM tooling or communities
- Familiarity with AI/ML tooling on AWS, Azure, or GCP