AI Native Transformation Manager

Accenture•Chicago, IL

19h•Hybrid

About The Position

Accenture is a global leader in AI and cloud transformation, partnering with leading cloud providers like Nvidia, AWS, Microsoft Azure, and Google Cloud to deliver end-to-end services. The Cloud Advisory Practice focuses on helping organizations define, plan, and implement innovative AI and cloud strategies that drive business value. This involves leveraging deep expertise across cloud platforms and technologies to design scalable, secure, and resilient cloud environments, offering guidance in areas such as agentic AI infrastructure & hosting, modern cloud foundation, security and resiliency, full-stack FinOps, and cloud-native development. The role is centered on solving complex enterprise AI transformation problems, including designing multi-agent systems, building composable architectures, and transforming established industries. The work combines AI technology, industry expertise, and entrepreneurial experience to fundamentally reimagine core business processes across financial services, healthcare, procurement, retail, and logistics. This is achieved by leveraging agentic architectures, multi-agent orchestration patterns, and composable AI systems alongside proven distributed patterns (Event Sourcing, Event-Driven Architecture, Microservices, Domain-Driven Design, CQRS) and technologies (Claude API, Neo4j, Qdrant, PostgreSQL, event streaming platforms, vector databases, cloud platforms) to build AI-native solutions. The approach, termed AI Transformation Decoupling, involves designing and building state-of-the-art agentic systems to wrap legacy cores, establish real-time feedback loops, add new AI-native functionality, and methodically transform existing systems into composable, event-driven architectures that support human-AI collaboration at scale. Before making changes, AI agents are used to systematically understand existing systems by mapping dependencies, analyzing git history, discovering hidden coupling, and identifying knowledge concentration. The team is described as deeply hands-on, highly technical, and battle-hardened, leading from the front as AI transformation thought leaders. Travel may be required for this role, varying from 0 to 100% depending on business need and client requirements.

Requirements

Minimum of 3 years of hands-on experience building interesting and innovative applications, with at least 1 year working with AI/LLM systems in production or production-like contexts.
Minimum of 3 years of experience explaining complex AI concepts to executive audiences and translating between technical capabilities and business value.
Minimum of 2 years of experience designing and building software systems, including planning AI-native architectures, infrastructure, and integration patterns.
Minimum of 5 years of experience leading an agile team and managing the unique challenges of AI development (iteration on prompts, dealing with non-determinism, managing costs).
Minimum of 1 year of experience designing engineering systems and DevOps for AI workloads (model deployment, monitoring, version control for prompts).
Minimum of 1 year of understanding of the economics of AI systems (token costs, latency tradeoffs, when to fine-tune vs. prompt).
Bachelor's degree or equivalent (minimum 12 years) work experience. (If Associate’s Degree, must have minimum 6 years work experience).

Nice To Haves

Experience building with LLM APIs (Claude, GPT-4, etc.) and understanding prompt engineering patterns.
Experience designing multi-agent systems with distinct roles (planning, execution, evaluation, coordination).
Experience designing and building MCP Server and Client including standard connection, tools and data exposure as well as specific tools.
Experience with agentic frameworks (LangChain, LlamaIndex, or custom orchestration patterns).
Experience analyzing and transforming existing systems—understanding legacy architectures through systematic dependency analysis, git history mining, and architectural discovery before modification (brownfield work is most of enterprise AI).
Hands-on experience with vector databases and RAG architectures (Qdrant, Pinecone, ChromaDB, Weaviate).
Understanding of graph databases and knowledge graphs (Neo4j, Neptune) for semantic relationships and ontology modeling.
Hands-on experience with cloud platforms (AWS, Azure, or GCP)—specifically choosing and configuring components for AI-native architectures.
Experience with event-driven architectures and streaming technologies (Kafka, Kinesis, EventBridge, event streaming platforms) for real-time AI feedback loops.
Experience with microservices architectures and composable system design.
Experience with containerization and orchestration (Docker, Kubernetes, ECS).
Understanding of observability and monitoring for AI systems (LLM tracing, token usage, latency, cost tracking).
Experience with production AI operations—LLMOps, prompt versioning, model lifecycle management, or managing AI systems at scale.
Experience with real-time communication protocols (WebSockets, Server-Sent Events, HTTP/2) for human-AI interaction patterns.
Experience with distributed transactional data stores and their consistency models.
Functional programming experience, particularly patterns relevant to AI systems (immutability, pure functions, composition).
Understanding of information retrieval, semantic search, or embedding-based similarity.
Prior experience in traditional ML/data science (helpful but not required—we're often doing something quite different).

Responsibilities

Design and build state-of-the-art agentic systems to wrap legacy cores.
Establish real-time feedback loops.
Add new AI-native functionality.
Methodically transform existing systems into composable, event-driven architectures that support human-AI collaboration at scale.
Use AI agents to systematically understand existing systems (mapping dependencies, analyzing git history, discovering hidden coupling, and identifying knowledge concentration).
Learn patterns proven in production and understand when each applies and why.
Build systems that scale because they're architecturally sound.