About this role
Accenture seeks a Full Stack AI Engineer Associate Director in Singapore to lead complex, multi-workstream agentic AI transformation programmes. The role spans client problem diagnosis, value hypotheses, solution architecture and production delivery, combining hands-on technical authority with senior stakeholder, team and account leadership.
Responsibilities
- Lead client workshops to identify agentic AI opportunities, define target operating models and develop transformation roadmaps. Quantify business cases, establish baselines and track outcomes in cost, cycle time and revenue.
- Architect and deliver production agentic systems, including agent harnesses, supervisor/worker and other orchestration patterns, A2A coordination, LLM gateways, tools and cloud infrastructure. Define knowledge layers using RAG, MCP-connected sources, Text-to-SQL, Elasticsearch, knowledge graphs and ontologies.
- Establish evaluation, AgentOps, LLMOps and DevOps practices covering golden datasets, LLM-as-judge, trajectory testing, CI/CD, deployment, registries, observability and drift detection. Set governance standards for guardrails, prompt injection defenses, agent identity, PII redaction, human approval, blast-radius controls and audit trails.
- Oversee concurrent workstreams and delivery quality; advise VP- and C-suite clients, lead governance forums, develop proposals and expansion opportunities, coach managers and engineers, and contribute reference architectures and practice development.
Core requirements: 3+ years building production agentic AI systems and 3+ years building production LLM applications; 8+ years in classical AI/ML, data engineering and advanced analytics; 10+ years in full-stack engineering and complex software delivery; 8+ years of cloud-native development on AWS, Azure or GCP; and 5+ years of technical leadership across multi-workstream programmes. Requires experience leading client-facing AI transformation, independently managing senior stakeholders, delivering significant digital transformation programmes, and hands-on command of agentic architecture, evaluation and safety. A bachelor's degree in a related field is required.
Additional strong signals: A master's degree in Computer Science, AI or Engineering; building evaluation disciplines; autonomy governance in regulated environments; commercial development; experience in financial services, healthcare or telecoms; multi-LLM orchestration and agent cost governance; integrations with SAP, Salesforce, ServiceNow or data lakes; published agentic AI work; and CXO-level technical advisory experience.