AI Application Developer

About this role

This is Cygnific’s first dedicated AI Developer role. You’ll be the hands-on builder in a newly forming AI capability, working directly with our AI & Emerging Technology Transformation Lead and alongside established RPA, Power Platform and UX colleagues.

That means real ownership: you’ll design, build, ship and keep running the AI solutions you create. There isn’t an AI engineering team or a mature AI platform to lean on yet. You’ll help build both.

Purpose of the role

You will design, build, test and run AI-enabled applications, agents and integrations that solve real business problems across Cygnific and our client environments. You’ll build primarily on Microsoft’s AI stack, while choosing the right tools, models and patterns for each problem.

The role complements Cygnific’s existing RPA and Power Platform capabilities. The goal is AI solutions that are useful, secure, responsible and valuable, and that make it from prototype into daily use.

What you’ll do

·       Build and maintain AI applications, copilots and agents, primarily with Azure OpenAI, Azure AI Foundry and Copilot Studio.

·       Design retrieval (RAG) solutions that ground AI answers in enterprise and client content, including chunking, embeddings, retrieval tuning and citations.

·       Design controlled agent workflows that call tools and APIs, follow business rules, involve people at the right decision points, and leave an auditable trail.

·       Build APIs, connectors and integrations into Microsoft 365, Teams, Power Platform, RPA, and client customer-service and CRM systems.

·       Create evaluation sets and regression checks for groundedness, accuracy and output quality, and use them before and after every release.

·       Monitor your solutions in production (latency, cost, errors, retrieval quality) and fix what breaks.

·       Apply data protection, access control and responsible AI standards, working with ICT, BI/Data and Governance colleagues.

·       Document architecture, data flows and configuration so solutions can be supported and reused.

·       Work with Innovation leads and UX to turn business problems into buildable solutions, and demo your work to technical and non-technical audiences.

·       Build reusable templates, components and patterns that make the next solution faster.

What you own, and where you’re supported

You own the application layer end to end: AI applications, agents and copilots; retrieval pipelines; APIs and integrations; testing and evaluation; monitoring and fixing your solutions in production; documentation.

You’re supported on cloud platform foundations (hosting, networking, identity and security operations). You’ll need to know what your solutions require from the platform and say so clearly. Building or operating the platform itself is outside this role.

What we’re looking for

Must have

1.       You have built and shipped at least one LLM-based application that internal users or clients relied on in their day-to-day work - a running solution you kept improving, beyond a demo or short pilot.

2.       You have built retrieval (RAG) against real business content, including chunking, embeddings, retrieval tuning, and grounding or citations

3.       You write strong Python or TypeScript and build APIs, services and integrations, beyond notebooks or no-code flows

4.       You have tested AI output systematically (evaluation sets, regression checks) and monitored a solution after it went live

5.       You explain technical trade-offs clearly to non-technical stakeholders, in English

Strong advantage

Hands-on experience with Azure OpenAI, Azure AI Foundry, Copilot Studio or Power Platform. Experience on another cloud is welcome if you can show a credible path to Azure.

Also valued

·       C#, and Azure services such as Key Vault, Storage, Azure SQL and Application Insights

·       Embedding AI into Power Apps, Power Automate and Dataverse; building custom connectors and webhooks; vector search and re-ranking

·       Practical experience with GDPR, data governance or EU AI Act requirements

·       Experience in agile teams

Experience

Typically 5+ years building software, including recent hands-on work on LLM-based applications. We weigh what you’ve built and shipped more heavily than years of experience or job titles.