Senior .NET developer with AI

Tech Stack:

• C#, .NET 8+, ASP.NET Core

• Azure, Azure OpenAI, Azure AI Search, LLM

• AI development tools: Claude Code, Codex

• CI/CD

• Bicep/Terraform

Project Description:

We are looking for a strong Senior .NET Backend Developer to lead the development of AI capabilities in a SaaS platform that delivers procedures and practice guidance to public sector and social care organisations in the UK. You will work closely with the client’s Head of Product, Engineering Manager and a mixed team of client and Spyrosoft engineers. The role combines hands-on development with technical leadership. You will shape the AI architecture and roadmap, build production-grade RAG and AI search features on Azure (Azure OpenAI, Azure AI Search, Semantic Kernel), integrate them with a headless CMS (Squidex), and make sure everything meets the quality, safety and compliance standards of a regulated domain. Knowledge transfer is a key part of the engagement, so the client’s team can own and run everything that is built.

Required Qualifications:

• 7+ years of professional C#/.NET development, including recent .NET (8 or later), ASP.NET Core and REST API design

• Hands-on experience shipping LLM-based features to production (ideally RAG or AI search), not only prototypes

• Working knowledge of embeddings, vector search, chunking strategies and prompt design, and how they affect answer quality

• Strong Azure experience, including Azure OpenAI or Azure AI Search, plus core services (App Service or AKS, Functions, Storage, Key Vault)

• Experience using AI tools to accelerate development (Claude Code, Codex etc.)

• Sound engineering habits: automated testing, CI/CD, observability and infrastructure as code (Bicep or Terraform)

• A clear, practical approach to data protection and security when handling sensitive information

• Ability to explain technical trade-offs plainly to product, compliance and customer-facing colleagues

• English level C1 or higher

Nice to Have:

• Experience with a headless CMS (Squidex, Contentful, Umbraco Heartcore, Kontent.ai or Strapi), including content modelling, APIs and webhooks; Squidex specifically (.NET client, GraphQL API, content workflows, self-hosting on Azure) Semantic Kernel, Kernel Memory or similar .NET AI orchestration frameworks LLM evaluation tooling and methods (e.g. Azure AI Foundry evaluations, RAGAS or custom test harnesses)

• Multi-tenant SaaS design, especially isolating content and data per customer MongoDB or Cosmos DB, and search relevance tuning

• Front-end experience (React or Blazor) to prototype AI user experiences

• Experience in public sector, health, social care or other regulated domains, ideally with familiarity with UK safeguarding and social care practice

Responsibilities :

• Agreeing the architecture and roadmap for AI capabilities with the Head of Product and Engineering Manager

• Leading design reviews, code reviews and technical decisions for AI work, documented as lightweight decision records

• Mentoring and upskilling C# developers in AI engineering practice, prompt design and evaluation

• Making build-versus-buy recommendations on models, services and tooling, with cost and risk set out clearly

• Building RAG for grounded search and question answering over procedures, with citations back to source

• Implementing hybrid search (keyword plus vector) and relevance tuning, including per-authority content scoping

• Delivering AI features such as summarisation, “what’s changed” highlighting and AI-assisted authoring for content editors Integrating with Squidex (or a similar headless CMS) and building ingestion pipelines that keep indexes in sync with published content, versioning and workflow states

• Designing and running AI workloads on Azure, together with infrastructure as code and CI/CD pipelines built with the SRE team

• Monitoring latency, quality and cost with Application Insights and Azure Monitor, keeping token spend predictable

• Building evaluation sets and automated checks for answer accuracy, grounding and regressions before each release

• Putting guardrails in place against hallucination, prompt injection and cross-tenant content leakage

• Ensuring compliance with UK GDPR, data residency and accessibility (WCAG 2.2 AA) requirements

• Taking AI features through Beta and GA quality gates with clear evidence

• Keeping architecture, runbooks and decision records current, and pairing with client engineers so the capability stays in-house after the engagement