Senior AI Software Engineer
Centrica plc24 hours ago
Windsor, CanadaSenior
Responsibilities
- Design and build agentic services on Amazon Bedrock and AgentCore Runtime, including multi-step orchestration, state and memory management, and failure handling.
- Build and own reliable tool and function-calling interfaces for capabilities such as account lookup, billing, and identity.
- Develop retrieval-augmented generation using Bedrock Knowledge Bases, including content preparation, chunking, embeddings, retrieval evaluation, and answer grounding.
- Version, test, and manage prompts as engineering artifacts.
- Define evaluation metrics and build regression harnesses for accuracy, containment, safety, and release decisions.
- Implement guardrails and safety controls covering PII, out-of-scope refusal, jailbreak resistance, and human escalation.
- Optimize real-time AI systems for latency and cost through streaming, model selection, routing, caching, and token budgets.
- Instrument and monitor agent behavior using Datadog and QuickSight, and support live services by investigating incidents and implementing permanent fixes.
- Document build-versus-buy decisions, write Architecture Decision Records, and take designs through AI governance and change-control processes.
- Track developments in generative AI and contribute suitable capabilities to the product roadmap.
Requirements
- Substantial software engineering experience with strong Python and production-code development skills.
- Demonstrable production delivery of LLM or agentic systems for real users.
- Hands-on experience with tool and function calling, multi-step agent orchestration, and conversational state management.
- Production experience with retrieval-augmented generation, retrieval-quality evaluation, and answer grounding.
- Experience building or owning evaluation harnesses, defining meaningful metrics, and using them for release decisions.
- Experience deploying and operating cloud AI services with serverless or container-based architectures, infrastructure as code such as Terraform, and CI/CD.
- Practical responsible-AI experience covering data privacy, guardrails, bias, explainability, and human-in-the-loop design under compliance obligations.
- Ability to debug non-deterministic systems and analyze behavior statistically.
- Strong collaboration and communication skills with product and governance stakeholders.
- Degree in Computer Science, Engineering, or a related discipline, or equivalent practical experience.
- Amazon Bedrock, AgentCore Runtime, Bedrock Knowledge Bases, and Bedrock Guardrails experience is desirable, though comparable foundation-model platform experience transfers.
- Voice AI, real-time streaming, speech-to-speech, barge-in, turn-taking, ASR, TTS, Amazon Connect, Model Context Protocol, fine-tuning, distillation, model routing, regulated-industry delivery, and AWS Associate-level certification are desirable.
Benefits
- UK-based hybrid work arrangement with occasional travel to site.
- Total rewards designed to provide flexibility and support for employees and families financially, physically, and emotionally.
- Opportunities for growth, development, and participation in sustainability-focused AI products.
Categories
About Centrica plc
Centrica plc supplies gas, electricity, and related services to homes and businesses, primarily in the UK and Ireland, through brands such as British Gas and Bord Gáis Energy. It also trades energy, manages storage and distributed energy assets, and provides home services and connected-home technology. Headquartered in Windsor, Berkshire, the company was created in 1997 from the demerger of British Gas and is listed on the London Stock Exchange.