Public SectorAI Agent

Camden Council

Alpha phase for agentic support

A narrowboat on Regent's Canal at Camden Lock, reflected in still water
Overview

Council staff needed answers from data spread across applications, databases and internal documentation.

Working with Aten Consultancy, Level designed and prototyped an AI agent that queries every source at once — so staff spend less time hunting for information and more time acting on it.

  • Conversational interface across apps, databases and documents

    1

  • Evaluation dimensions in LangSmith — tool use, cost, speed, model performance

    4

  • Infrastructure as code — Azure deployed via Bicep

    100%

Our Role

We delivered a fully working prototype: one conversational interface that lets council staff query and analyse data across disparate systems.

  • Internal LLM chat interface for council employees
  • RAG and agent tooling for dynamic data retrieval
  • LLM integration using Azure Foundry models
  • Evaluation in LangSmith: tool use, cost, speed and model performance
  • Azure infrastructure deployed as code with Bicep
  • Future-ready architecture for scaling AI across the council
Explore more work

Proud of the work we do,
and the people we help.

Show me more work!