Project Role : Large Language Model Architect
Project Role Description : Architect large language models (LLM) that can process and generate natural language. Design neural network parameters, trained on large quantities of unlabeled text data.
Must have skills : Large Language Models (LLMs)
Good to have skills : NA
Minimum 18 year(s) of experience is required
Educational Qualification : 15 years full time education
ROLE OVERVIEW
The Lead Architect is the senior-most technical authority on Accenture's internal Agentic AI Platform — the SDK and runtime layer that sits above the enterprise tool stack and below every AI agent the company builds. This is not an advisory role. You will define and own the design contracts across context and ontology management, agent lifecycle, and governed tool integration, and you will hold those boundaries against pressure from IT, business stakeholders, and the platform's own engineering pods. Your decisions will be load-bearing for years.
KEY RESPONSIBILITIES
Own the end-to-end technical architecture of the platform SDK — context management, agent lifecycle, tool registry, and their integration contracts
Set and enforce design standards across six engineering pods building in parallel under two Engineering Managers
Make technology selection decisions for the orchestration framework, knowledge graph backend, and LLM runtime — with full awareness of the enterprise procurement and IT landscape
Define the ontology schema governance model and arbitrate cross-domain schema conflicts with Finance, Legal, Procurement, and HR
Lead the technical relationship with Jason (enterprise architect, CIO team) — the most critical external technical dependency for the platform
Translate CISO and compliance requirements into implementable platform specifications for the Security and Governance pod
Review and approve all cross-pod architecture decisions own the platform's technical debt register
IDEAL PROFILE
Has been the principal technical decision-maker on at least one platform or developer infrastructure product that other engineers built on top of — not a contributor, the owner
Transitioned into AI and LLM systems engineering early enough to carry real production experience, not credentials 3–5 years of hands-on LLM systems work minimum
Came up through distributed systems, data infrastructure, or enterprise architecture — the kind of background that produces strong opinions about reliability, API design, and failure modes
Comfortable in large enterprise organizations: understands that architecture decisions are also organizational decisions, and has navigated IT, security, and business stakeholder dynamics before
Does not need the architect title to hold authority — can hold a technical boundary through reasoning, not hierarchy
15 years full time education
