AI Tech Lead

Technical Leadership & Architecture

• Own the technical delivery of AI solutions from ideation through to production deployment
• Define and enforce solution architecture patterns including single-agent, multi-agent, and orchestration topologies
• Make technology selection decisions with cost, performance, and maintainability trade-offs
• Design and approve prompt architectures, retrieval strategies, and context assembly patterns
• Establish reference architectures and reusable design patterns for the AI engineering team

Delivery Management

• Lead AI projects through a stage-gated delivery model: Ideation → Framing → Design → Build → Test → UAT → Production
• Own design approval gates and ensure every deliverable meets agreed quality bars before stage exit
• Manage non-functional requirements (latency, cost-per-task, throughput) and ensure they are met at each gate
• Drive parallel workstreams (e.g., systems data readiness alongside agent build) to prevent schedule slip
• Coordinate cross-team dependencies with platform, MLOps, data, and business teams

Team Leadership

• Lead, mentor, and grow a team of AI Engineers across agent development, prompt engineering, and LLM integration
• Conduct code reviews and AI engineering quality checks
• Establish and enforce development standards, coding guidelines, and engineering best practices
• Foster a culture of experimentation, continuous learning, and knowledge sharing
• Manage team capacity, sprint planning, and delivery commitments

Agentic AI & LLM Expertise

• Design and oversee agentic orchestration solutions (multi-agent systems, state management, human-in-the-loop patterns)
• Lead prompt engineering strategy including system prompt design, few-shot patterns, chain-of-thought reasoning, and prompt versioning
• Architect Retrieval-Augmented Generation (RAG) pipelines including chunking strategies, embedding models, and retrieval quality evaluation
• Drive evaluation harness design: golden-set creation, grading rubrics, pass-bar thresholds, and CI-integrated eval suites
• Ensure non-determinism is managed through stable eval scores across consecutive runs

Security & Governance

• Ensure all AI solutions follow zero-trust security principles: managed identities, no secrets in code, defence in depth
• Integrate content safety checks on all LLM inputs and outputs
• Design and enforce guardrails for agent actions including blast-radius assessment and fallback paths
• Own adversarial testing strategy: prompt injection, jailbreak, and tool-misuse test packs
• Contribute to governance frameworks including model cards, explainability documentation, and audit evidence

Stakeholder Engagement

• Translate business requirements into feasible AI solution designs
• Present architecture decisions, PoC findings, and delivery progress to senior stakeholders
• Collaborate with business SMEs on use-case framing, autonomy-level decisions, and acceptance criteria
• Participate in joint go/no-go gates for production deployment

Observability & Operations

• Define observability and trace design requirements for AI workloads (logging, telemetry, cost tracking)
• Establish alert thresholds, on-call procedures, and hypercare plans for production AI systems
• Drive token usage monitoring, cost optimisation, and model-tier selection strategies
  • Own post-deployment monitoring, prompt optimisation, and continuous improvement cycles