What you’ll do
Build the backend systems our AI capabilities run on: services and APIs first, with LLM features, retrieval and cloud infrastructure around them. You will work across backend services, AI-powered features, data pipelines, search and RAG systems, and cloud-native deployment.
Day to day:
- Design, develop and maintain scalable, secure, high-performance backend services and APIs.
- Build and integrate LLM-powered applications: RAG, semantic search, document processing and AI agents.
- Work with vector databases and vector search for embedding-based retrieval.
- Design and optimise backend architectures for cloud-native, distributed systems.
- Develop integrations with cloud services, APIs, databases, storage systems and AI platforms.
- Build reliable ingestion and processing pipelines for AI and search workloads.
- Implement authentication, authorisation, observability, logging and monitoring.
- Collaborate with frontend, AI/ML, DevOps and product teams to deliver end to end.
- Contribute to architecture decisions, code review, testing and CI/CD.
- Optimise performance, scalability, reliability and cloud resource use.
We look for
Someone who goes beyond calling an LLM API. You understand how to build production systems around AI capabilities — ingestion, document processing, embeddings, vector retrieval, APIs, security, scalability, deployment and monitoring — and you are comfortable moving from backend architecture to AI and search integration to cloud infrastructure and Kubernetes.
Required:
- Three years or more of professional backend development.
- Strong proficiency in a backend language: Python, TypeScript/Node.js, Java, Go or C#.
- REST APIs, microservices, distributed systems, databases and backend architecture.
- Hands-on work with LLMs and LLM-based applications.
- Practical experience with RAG, embeddings, semantic search or vector search.
- Vector databases and search technologies: Azure AI Search, Qdrant, Pinecone, Weaviate, pgvector, Elasticsearch/OpenSearch or similar.
- One major cloud platform, preferably Microsoft Azure.
- Docker and Kubernetes.
- Clean code, testing and system design, and the judgement to debug complex production issues.
Nice to have:
- Azure OpenAI, Azure AI Search, or Azure AI Foundry.
- Production RAG pipelines and document intelligence systems.
- Embedding models, chunking strategies, retrieval optimisation, reranking and RAG evaluation.
- Redis, Kafka, RabbitMQ or similar messaging and caching.
- Microservices architectures and infrastructure as code.
- Observability with Prometheus, Grafana or OpenTelemetry.
- Agent frameworks: LangGraph, Microsoft Agent Framework, AutoGen or similar.
- Enterprise AI platforms, SaaS products or multi-tenant architectures.