AI Delivery Architect in London

Job Type: Permanent

Location: London

Number of hours: 40 hours per week – full time

Ready to utilize your experience and expertise in AI Delivery Architect? We have an exciting role for you!

Role Summary

We are looking for an experienced AI Delivery Architect to design enterprise AI solutions, validate their implementation and work with engineering teams to drive execution. This is a hands‑on delivery role, not an advisory‑only architecture position. The architect converts business objectives and domain inputs into an implementable solution, makes the key technical decisions, guides the engineering team and remains accountable for the technical integrity of the delivery.

The role is important because complex AI programs bring uncertainty across models, data, integration, security, evaluation and operations. The architect provides one coherent design, makes trade‑offs visible, reduces technical risk and ensures that the resulting solution is secure, reusable, observable and ready for production.

Pre-requisites

• Technology experience, including architecture and technical delivery of enterprise solutions and recent hands‑on experience with AI‑enabled systems beyond PoC
• Strong solution architecture foundation across applications, data, APIs and integration, cloud, security, deployment and operations
• Hands‑on understanding of GenAI and agentic AI, including LLMs, RAG, context engineering, tool use, orchestration, memory and human‑in‑the‑loop (HITL) controls
• Experience designing production‑grade integrations with enterprise systems using APIs, microservices, events, streams or queues
• Experience defining evaluation, testing, observability, security, responsible AI, reliability, performance and cost controls for AI solutions
• Practical experience with at least one major cloud platform and its AI, data, identity, networking, monitoring and deployment services
• Ability to lead multidisciplinary delivery teams, review critical implementation, communicate technical trade‑offs and transfer ownership to the wider delivery organization

Key Responsibilities

• Design the end‑to‑end AI solution across application, model, data, context, integration, security, deployment, evaluation, observability and operations
• Translate business objectives and domain‑team inputs into solution boundaries, architecture decisions, technical roadmaps, NFRs and measurable acceptance criteria
• Select appropriate AI patterns, models, platforms and implementation approaches based on quality, safety, latency, cost, data sensitivity, maintainability and operational risk
• Define architecture for AI agents, orchestration, tool invocation, memory, retrieval, model abstraction, governed execution and human escalation
• Define the data and knowledge approach, including ingestion, metadata, lineage, access‑aware retrieval, ranking, freshness and evidence traceability
• Define reusable integration patterns for APIs, MCP, microservices, events, enterprise systems and legacy platforms, including authentication, authorization, idempotency and failure handling
• Define standards for model gateways, routing, prompt and configuration versioning, environment promotion, release controls and rollback
• Define the evaluation strategy, test coverage, quality thresholds and release gates across retrieval, prompts, models, agents, tools and end‑to‑end outcomes
• Define end‑to‑end observability covering model calls, retrieval, orchestration, tools, data pipelines, latency, usage, cost, errors and audit events
• Define production‑readiness requirements for scalability, availability, resilience, DR, security, privacy, responsible AI and ongoing support
• Validate implementation through architecture reviews, critical code and design reviews, technical walkthroughs and production‑readiness checks
• Resolve cross‑team technical risks and work with engineers to ensure the agreed design is implemented effectively
• Create reusable reference architectures, ADRs, interface contracts, engineering patterns, playbooks and operational guidance
• Coach engineers through paired design and reviews, and progressively transfer architecture and operational knowledge to the wider team

Required Skills & Qualifications

• Architecture & Delivery Leadership
• Solution and enterprise architecture across cloud‑native, hybrid and legacy environments
• Architecture governance, technical design authority, ADRs and NFRs
• Ability to convert ambiguous requirements into an executable design and phased delivery plan
• Ability to review implementation and diagnose issues across application, data, integration and platform layers
• LLMs, structured outputs, prompt and configuration management, grounding and common failure modes
• Agentic patterns covering planning, routing, delegation, tools, state, memory, durable workflows and human approval
• RAG, semantic or hybrid search, vector databases, metadata, knowledge graphs and access‑aware retrieval
• Data ingestion, data quality, metadata, lineage and schema evolution
• Cloud services, containers, Kubernetes, IaC, CI/CD, identity, secrets, networking, logging and monitoring
• MLOps or LLMOps practices covering versioning, deployment, evaluation, monitoring and rollback

AI Quality, Governance & Operations

• Evaluation of groundedness, task success, policy adherence, tool selection, safety, latency and cost
• Responsible AI, privacy, security, explainability, auditability and human accountability
• Operational resilience, performance, capacity, incident response and production‑support readiness
• Metrics, logs, traces, dashboards and alerts across the complete AI service

Leadership & Professional Skills

• Clear communication with business, architecture, engineering, security, risk and operations stakeholders
• Decisive, evidence‑based judgement in ambiguous delivery situations
• Ability to surface risks and limitations without slowing execution unnecessarily
• Strong collaboration, mentoring and influencing skills across distributed teams
• Balanced focus on delivery outcome, engineering quality, cost, risk and maintainability

Preferred Qualifications

• Architecture leadership in regulated, security‑sensitive or high‑availability environments
• Experience creating reusable AI platform patterns, agent control planes or engineering standards
• Experience with multi‑model, hybrid, on‑premises or sovereign AI environments
• Relevant cloud solution architect certification from AWS, Microsoft Azure or Google Cloud
• Relevant AI, ML, data, security, Kubernetes or service‑management certification
• Bachelor’s or master’s degree in Computer Science, Engineering, Data Science, Mathematics or a related field, or equivalent professional experience

What Success Looks Like

• The solution design is implementable and guides delivery rather than remaining conceptual
• Implementation conforms to the agreed architecture and reaches production with appropriate quality and controls
• Reusable patterns reduce effort and technical risk for subsequent AI use cases
• The wider delivery team can independently extend, operate and govern the solution after the incubation phase

TCS is consistently voted a Top Employer in the UK and globally. Our competitive salary packages feature pension, health care, life assurance, laptop, phone, access to extensive training resources and discounts within the larger Tata network.

We offer health & wellness initiatives and sports events; we are the proud sponsor of the London Marathon.

Diversity, Inclusion and Wellbeing

Tata Consultancy Services UK&I is committed to meeting the accessibility needs of all individuals in accordance with the UK Equality Act 2010 and the UK Human Rights Act 1998. We welcome and embrace diversity in race, nationality, ethnicity, disability, neurodiversity, gender identity, age, physical ability, gender reassignment, sexual orientation. We are a disability inclusive employer and encourage disabled people to apply for this role.

As a Disability Confident Employer, we offer an interview to applicants with disabilities or long‑term conditions who meet the minimum criteria for the role.