Senior GenAI and Agentic Developer - Python - Fastapi - AWS Bedrock - Llms

Senior GenAI & Agentic Developer | Python | FastAPI | AWS Bedrock | LLMs | London Hybrid Salary: £75,000 - £87,000 + bonus + benefits Contract: Permanent Location: London - hybrid, 2-3 days a week on client site Eligibility: UK based with existing right to work - must hold a dependent visa as a minimum requirement

Requirement: Production Python (FastAPI) and hands-on LLM integration on AWS (Non-Negotiable) 🚨 Please only apply if you have commercial experience with ALL of the following 🚨

• Strong Python backend engineering in production with FastAPI
• Building Generative or Agentic AI features with LLMs into real apps or workflows
• AWS services including EC2, EKS, S3, SQS, Kinesis/Firehose and Bedrock
• Evaluating LLM performance and prompt handling complexity
• CI/CD and observability in regulated enterprise environments Optimal have partnered with a global next-gen engineering consultancy delivering AI, cloud and automation programmes for some of the world's biggest organisations. You'll build core services on an enterprise-grade Generative AI platform for a major global insurer - real production AI, at serious scale, in a regulated environment where getting it right matters. This is a senior hands-on engineering role. You'll build the services, APIs and GenAI components that teams across the business rely on, and make sure everything that ships is secure, observable and built to last. Required Background
• Senior-level Python backend engineering in production systems
• Practical experience integrating LLMs into enterprise systems
• Experience evaluating LLM performance, cost and prompt handling complexity
• Solid DevOps mindset with CI/CD and observability expertise
• Strong grasp of authentication, secret management, network boundaries and model access governance ℹ️ Very Important Notes
• This is hands-on - not suitable for architects or managers who have stepped away from code
• Python backend depth is essential - frontend-heavy or data science-only profiles won't be considered
• Must be comfortable working under strict enterprise security controls
• 2-3 days a week on client site in London Must-Haves
• Generative and Agentic AI patterns: LLM integration, prompt orchestration, output validation and evaluation tooling
• CI/CD expertise, ideally Azure DevOps
• Observability built in from day one
• AWS: EC2, EKS, S3, SQS, Kinesis/Firehose, Bedrock and AgentCore
• Proactive self-starter - spots gaps and drives solutions without waiting for tickets Bonus Experience
• Kong API Gateway, Kong Mesh, Flux CD
• IaC and Terraform with GitOps workflows
• Prompt red teaming and evaluation tools (Promptfoo, Arize)
• SQL (MySQL, PostgreSQL) and NoSQL (MongoDB, Cassandra) databases
• RAG patterns and vector search technologies Hands-On Experience With
• Python (FastAPI) backend services and APIs
• LLMs, RAG and agent systems
• AWS Bedrock and AgentCore
• Azure DevOps CI/CD pipelines
• Microservices and RESTful APIs What You'll Be Doing GenAI Engineering
• Build and maintain backend services and APIs enabling secure access to LLM capabilities
• Develop and optimise Python GenAI and Agentic components: prompt orchestration, output validation and evaluation tooling
• Integrate LLMs with enterprise systems, observability and security frameworks
• Contribute to RAG implementations and data access patterns Quality & Performance
• Drive benchmarking, evaluation and experiment tracking for LLM performance and cost
• Troubleshoot and optimise performance across distributed systems and cloud services
• Design and maintain CI/CD pipelines (Azure DevOps primarily) Collaboration
• Partner with platform leads, architects and SRE teams to operationalise GenAI services reliably
• Shape API standards, reusable patterns and documentation for platform adoption
• Make the platform something engineering teams genuinely love using What They're Looking For
• A senior hands-on engineer who owns what they build
• Someone comfortable in emerging domains where best practice is still being written
• A collaborative communicator who earns trust across engineering, data and product
• An engineer who adopts new GenAI techniques in a structured, secure way
• A bias toward automating everything If you tick every box above and want to build production GenAI at enterprise scale, get in touch for a fast response.