About this role
Salary: £60,000 - 100,000 per year
Requirements: We are looking for substantial software engineering experience, with strong Python and production code experience.We need demonstrable delivery of LLM or agentic systems into production for real users.We need hands-on experience with tool and function calling, multi-step agent orchestration, and managing conversational state across turns.We need production experience with retrieval augmented generation, including retrieval quality evaluation and answer grounding.We need a rigorous approach to evaluation, including building or owning an eval harness, defining relevant metrics, and using them to make release decisions.We need experience deploying and operating AI services in the cloud using serverless or container-based architectures, infrastructure as code such as Terraform, and CI/CD.We need practical application of responsible AI, covering data privacy, guardrails, bias, explainability, and human-in-the-loop design in a regulated environment.We need the ability to debug non-deterministic systems and reason about behaviour statistically.We need strong collaboration and communication skills.We need a degree in Computer Science, Engineering or a related discipline, or equivalent practical experience.Desirable: experience with Amazon Bedrock, including AgentCore Runtime, Bedrock Knowledge Bases and Bedrock Guardrails.Desirable: voice AI experience, including speech-to-speech models, real-time streaming, barge-in and turn-taking, or ASR and TTS pipelines.Desirable: any contact centre exposure, including Amazon Connect.Desirable: familiarity with agent interoperability standards such as the Model Context Protocol.Desirable: experience with fine tuning, distillation or model routing where it materially improved cost or latency.Desirable: delivery in a regulated industry, including working with vulnerable customer requirements or comparable obligations.Desirable: AWS certification at Associate level or above.We need core competencies in designing, integrating and operating AI-enabled solutions within enterprise environments, including prompt-driven workflows, retrieval-augmented systems and AI agents.We need structured evaluation, testing and monitoring practices to ensure AI outputs are reliable, secure and compliant with organisational guardrails.We need experience preparing and managing data used in AI workflows and taking responsibility for the responsible lifecycle of AI features from experimentation through deployment and continuous improvement.We need safe and responsible use of AI tools, with clear knowledge of when AI use is appropriate and strong awareness of accuracy, bias and compliance.We need the ability to design and reuse prompt templates to support consistent, high-quality workflow outputs, and to use AI to triage, classify and analyse information within our policy guardrails.We need the ability to recognise higher-risk scenarios and escalate to governance or security as needed.We need proficiency in enterprise AI co-pilots, knowledge assistants and AI-enhanced productivity tools. Responsibilities: We will have you join our Channel Platforms team, which owns our customer conversation estate across voice, chat and email.You will own the intelligence layer of our AI products, including reasoning and orchestration, tools, knowledge retrieval, guardrails and evaluation.You will design and build agentic services on Amazon Bedrock and AgentCore Runtime, including multi-step orchestration, state and memory management, and graceful handling of failure and ambiguity.You will build and own the tool layer that agents act through, exposing business capabilities such as account lookup, billing and identity as reliable, well-described actions with sensible error semantics.You will build retrieval augmented generation over Bedrock Knowledge Bases, owning content preparation, chunking, embedding strategy, retrieval quality and answer grounding.You will treat prompts as engineering artefacts by versioning, testing and understanding what changes when behaviour shifts.You will build the evaluation harness, define accuracy, containment and safe behaviour metrics, assemble regression sets, and enable evidence-based release decisions.You will implement guardrails and safety controls, including PII handling, out-of-scope refusal, jailbreak resistance and clean escalation to a human advisor.You will engineer for latency and cost, including streaming, model selection and routing, caching and token budgets.You will instrument agent behaviour end to end so that tool failures, retrieval misses and abandoned conversations are visible in Datadog and QuickSight.You will support live AI services in production, investigate incidents, diagnose non-deterministic behaviour, and close the loop with permanent fixes.You will make and document build-versus-buy decisions, write Architecture Decision Records, and take designs through our AI governance board and change control process.You will track developments in the generative AI landscape and bring the parts that deliver clear benefit into our roadmap, with an honest view of risk and upside. Technologies: AIAI AgentsAWSCI/CDCloudDatadogSupportLLMPythonSecurityServerlessTerraformProduct Owner More:
We are a family of brands revolutionising how we power the planet, with around 21,000 colleagues working together to create a greener, fairer future and a cleaner energy system that does not rely on fossil fuels. We do energy differently: we make it, store it, move it, sell it and mend it. This is a UK-based hybrid role with occasional travel to site. We offer total rewards designed to support different realities and help our people and their families financially, physically and emotionally. We are a people place, and we want you to find purpose, growth and a team where your voice matters.
last updated 38 week of 2026