Free with a WorkJio account: we compare your profile with this job and show what you already meet.
Share with a friend
Start
Not stated
Work mode
On-site
Experience
Some experience
About the role
This role leads technical delivery of agentic AI programmes while staying hands-on as an engineer. You will work directly with clients to understand their needs, turn those into agentic application designs, and guide a team of engineers to ship production-grade systems that deliver measurable value. It suits someone with real technical depth who is equally credible in a code review and in a client workshop.
What you'll do
Architect and deliver agentic AI solutions end-to-end, including agents, orchestration, tool layers, knowledge pipelines and full stack applications, contributing directly to code and design.
Design agent harnesses, orchestration topologies, A2A coordination patterns and LLM gateway configuration across providers.
Build and maintain MCP servers, translate business processes into agent skills and reusable workflows, and implement knowledge layer components such as RAG, Text-to-SQL, Elasticsearch and knowledge graphs.
Full Stack AI Engineer job at Accenture, Singapore | WorkJio
Set prompt architecture standards covering versioning, A/B testing and structured output schemas, and apply reasoning patterns like ReAct, CoT and ToT to each use case.
Build and operate evaluation and AgentOps pipelines, including golden datasets, LLM-as-judge, trajectory evaluation, agent testing suites, CI/CD for agents and prompts, asset registry management, production observability and drift detection.
Implement trust, safety and governance components such as guardrails, prompt injection defences, agent identity scoping, PII redaction, blast radius controls, HITL approval gates and audit trail design.
Act as the primary technical contact for client stakeholders, running workshops, translating business requirements into agent solution designs and explaining technical decisions to non-technical audiences.
Develop initial value hypotheses for agentic solutions, identifying automation and augmentation opportunities, estimating business impact and setting baseline metrics.
Contribute to solution design and proposals, spot expansion opportunities within current engagements and support account growth.
Lead workstream delivery in agile settings, managing scope, quality, technical risk and milestone accountability.
Establish DevOps, AgentOps and LLMOps practices including CI/CD, automated evaluation gates, deployment strategies and production operations.
Lead, manage and develop a team of Consultants and Analysts through clear expectations, technical coaching and structured code reviews and design sessions.
Foster a delivery culture of engineering rigour, continuous improvement and learning.
Keep hands-on knowledge of agentic AI developments current by testing new frameworks, tooling and research and bringing relevant advances into practice.
Contribute to internal practice development through reusable accelerators, reference implementations and delivery standards.
Good to know
Full time role based in Singapore.
Bachelor's degree in a related field required; a Master's degree is highl
Requirements
Experience building LLM-based applications in production with operational accountability
Experience designing and delivering agentic AI systems with meaningful autonomy in production
Hands-on experience with at least one agent orchestration framework (LangGraph, AutoGen, CrewAI, AWS Strands or equivalent) in production
Experience across the Agent Development Lifecycle: specification, harness build, tool and MCP integration, evaluation, deployment, observability and refinement
Track record deploying software systems in production with measurable reliability, performance or business value results
Experience with knowledge layer engineering: RAG pipelines, MCP-connected sources, Text-to-SQL, Elasticsearch and knowledge graph design
Apply on the company website
You will leave WorkJio. Never pay a fee or share your Singpass password, OTP or bank log-ins to apply.
