Director, AI Software Engineering (India/China)

This role could be based in India and China. We seek a senior software engineering leader to build and run the engineering backbone of our AI Center of Excellence (AI CoE). This role owns the end-to-end (E2E) delivery of Python-based AI platforms and products, from design and secure development through model/service deployment, production operations, governance, and lifecycle management in a regulated banking environment. You will lead multi-disciplinary team members to turn prototypes into compliant, scalable, reliable systems that create measurable business value.

Key Duties

• Leadership & Strategy
  • Define the AI engineering vision, roadmap, and operating model for the AI CoE; translate business priorities into executable, value-driven delivery plans.
  • Establish standards for AI software engineering excellence: secure SDLC, architecture review, technical design, coding conventions, documentation, and inner-source practices.
  • Drive a product mindset: clear backlogs, fast iteration cycles, and outcome-based KPIs.
• Platform & Architecture (Python Full-Stack)
  • Architect and deliver reusable Python platforms (backend), ReAct/Vue (frontend) and reference services for: LLM, RAG/GraphRAG services, agent orchestration, prompt/version management, evaluation-as-code, guardrails, and content moderation.
  • Data engineering, feature/vector stores; retrieval pipelines; Text-to-SQL and tool-calling services.
  • Own API and event-driven architectures (REST/gRPC/FastAPI; Kafka/PubSub), microservices, and front-end integration (React/TypeScript/Next.js) for internal users (RMs, analysts) and operational teams.
  • Ensure non-functional excellence: performance, cost, resiliency, observability, and operability (SLOs, error budgets).
• E2E Productionization in Banking
  • Lead the path from PoC MVP production, including change management, release governance, DR/BCP, capacity planning, and runbooks.
  • Implement robust CI/CD & GitOps (build/test/scan/sign/release) in Azure DevOps for programming services, models, and data pipelines across hybrid infra (Kubernetes/OpenShift; on-prem & cloud).
  • Partner with model risk and validation to embed evaluation gates (offline/online AB tests, drift/guardrail monitors) before and after go-live.
• Security, Risk & Compliance
  • Ensure data privacy, secrets, access control, and lineage; enforce SAST/DAST/SCA, container/image signing, SBOMs, data contracts, and audit trails.
  • Align delivery with bank governance (model risk, data privacy, records retention, third‑party risk); streamline approvals with evidence‑backed automation.
  • Champion safe‑by‑design AI: red‑teaming, prompt‑injection defenses, eval suites, and incident playbooks.
• Operations & SRE for AI
  • Stand up AI‑SRE practices: golden signals, OpenTelemetry instrumentation, autoscaling, canary/blue‑green deploys, shadow traffic, and rollback strategies.
  • Establish runtime cost/latency optimization (token budgets, caching, distillation/quantization, routing).
• People & Stakeholder Management
  • Build and mentor high‑performing teams (managers, leads, ICs) across backend, data/ML, and front‑end disciplines.
  • Partner with COE CPO, Data science team, CDO, AI platform, Security, Cloud, Data Platform, and Business Lines to unblock delivery; manage vendors/open‑source responsibly.

Key Responsibilities

• Strategy
  • Awareness and understanding of the Group's AI COE strategy and model appropriate to the role.
• Business
  • Awareness and understanding of how the AI models operating in the Group.
• Processes
  • Responsible for developing some Machine Learning/GenAI models.
• Risk Management
  • The ability to interpret the Machine Learning models, identify key issues and put in place appropriate model controls and measures.
• Governance
  • Identify key Responsible AI related to the models developed.
• Regulatory & Business Conduct
  • Display exemplary conduct and live by the Group’s Values and Code of Conduct.
  • Take personal responsibility for embedding the highest standards of ethics, including regulatory and business conduct, across the company. This includes understanding and ensuring compliance with, in letter and spirit, all applicable laws, regulations, guidelines and the Group Code of Conduct.
  • Effectively and collaboratively identify, elevate, mitigate and resolve risk, conduct and compliance matters.

Key stakeholders

• CIB AI COE

Skills and Experience

• Data Science (large) Data processing
• Data Analytics
• Programming skills, e.g., Python
• ML/DL frameworks, e.g., Scikit-learn, XGboost, Torch, Tensorflow
• Cloud Production deployment
• API Software risk
• AI Model Risk

Qualifications

• Bachelor’s degree (or above) in Computer Science, Computing engineering, Mathematics, or a related field.

Languages

• English

What we offer

• In line with our Fair Pay Charter, we offer a competitive salary and benefits to support your mental, physical, financial and social wellbeing.
• Core bank funding for retirement savings, medical and life insurance, with flexible and voluntary benefits available in some locations.
• Time‑off including annual leave, parental/maternity (20 weeks), sabbatical (12 months maximum) and volunteering leave (3 days), along with minimum global standards for annual and public holiday, which is combined to 30 days minimum.
• Flexible working options based around home and office locations, with flexible working patterns.
• Proactive wellbeing support through Unmind, a market‑leading digital wellbeing platform, development courses for resilience and other human skills, global Employee Assistance Programme, sick leave, mental health first‑aiders and all sorts of self‑help toolkits.
• A continuous learning culture to support your growth, with opportunities to reskill and upskill and access to physical, virtual and digital learning.
• Being part of an inclusive and values‑driven organisation, one that embraces and celebrates our unique diversity, across our teams, business functions and geographies - everyone feels respected and can realise their full potential.