Agentic Architect

About the role

Accenture Technology is looking for an Agentic Architect to design scalable, reusable AI agents that automate software development lifecycle (SDLC) and business processes. You will translate complex workflows into efficient multi‑agent solutions, leveraging cutting‑edge AI frameworks and cloud platforms.

Key responsibilities

• Map and document SDLC and business workflows for automation and agentification.
• Define agent typologies, interfaces, contracts, and maturity criteria.
• Design knowledge‑graph strategies, multi‑hop query capabilities, and memory architectures.
• Build and maintain a reusable agent catalog with version control.
• Identify high‑impact automation opportunities and define integration standards with DevOps tools (GitHub, Jira, CI/CD).
• Implement Human‑in‑the‑Loop mechanisms, quality gates, and observability (logging, OpenTelemetry).
• Collaborate with security teams to establish governance and guardrails.
• Promote AI‑native practices across development and business teams.

Required profile

• Minimum 3 years of experience designing multi‑agent systems or complex process automation.
• Strong understanding of modern SDLC processes (Agile/Scrum, CI/CD, testing, code review).
• Proven ability to model business processes and translate functional requirements into scalable agent architectures.
• Analytical mindset with solid problem‑solving skills.

Required skills

• Python programming.
• Agent frameworks such as LangChain, LangGraph, Semantic Kernel, AutoGen, BMAD or equivalents.
• Agentic Retrieval‑Augmented Generation (RAG) and Knowledge Graphs.
• Azure AI Foundry, Azure OpenAI or comparable cloud AI platforms.
• Experience with CI/CD pipelines, GitHub, Jira, and observability tools like OpenTelemetry.