GenAI Architect (AI&Data)

Job description

WHO WE ARE Accenture is a leading global professional services company that helps the world’s leading businesses, governments, and other organizations build their digital core, optimize their operations, accelerate revenue growth, and enhance citizen services. We offer solutions and assets across Strategy & Consulting, Technology, Operations, Industry X, and Accenture Song. At the AI&Data, we work locally and globally with over 250 Poland‑based specialists. Our expertise spans data architecture and sectoral specialization. We collaborate with tech partners like Microsoft, AWS, Google, Oracle, and Snowflake to implement cutting‑edge solutions. We provide comprehensive support from strategy definition to solution implementation and development, covering all aspects of Data & Analytics. THE WORK As an GenAI Architect, you will define how autonomous agentic AI ecosystems are designed, built, and governed at enterprise scale - systems where hundreds of specialized agents collaborate to solve real business problems for some of the most recognizable companies in the world. You will: Define enterprise AI strategy - deciding when to leverage commercial models vs. self-hosted open-source models, optimizing for cost, performance, security, and compliance Architect large-scale autonomous multi-agent ecosystems: designing agent orchestration, task delegation, context management, and agent-to-agent communication patterns across the organization Design the enterprise integration layer that connects autonomous agents to corporate systems - REST/API, event-driven architectures, MCP, and custom connectors - enabling agents to act within the existing technology landscape, not alongside it Design distributed vector architectures and knowledge layers: embeddings pipelines, indexing strategies, metadata management, and RAG policies that give agents access to enterprise knowledge at scale Own security, compliance, and quality at the architecture level: PII handling, authorization models, access control, agent quality metrics (factuality, hallucination rate, latency, cost per query), and ensuring adherence to regulations including the AI Act Own production excellence: enterprise-grade cloud architecture (AWS, Azure, GCP) with full observability, automated evaluation pipelines, monitoring, cost governance, and Responsible AI guardrails Act as a technical authority - setting architectural standards, coordinating across DevOps, Security, and business stakeholders, mentoring senior engineers, and shaping Accenture's global reference architecture for agentic AI Interview Note: During your meeting with our Technical Partners, we'd like to see how you approach complex architectural challenges in practice - using your preferred development environment, including any AI-assisted tools you work with. Be ready to share your screen and walk us through your thinking. WHAT’S IN IT FOR YOU? Direct impact on Accenture's global reference architecture for agentic AI - shaping how enterprise-scale autonomous systems are designed and governed. A platform as a recognized thought leader in the enterprise AI space - with visibility across global projects and C-level stakeholders. Work at the intersection of cutting-edge AI and real enterprise complexity - the problems you solve here don't exist anywhere else at this scale. HERE’S WHAT YOU’LL NEED Extensive experience (10+ years) in software architecture, data engineering, or ML engineering, with deep hands-on expertise in the GenAI and agentic AI space Proven track record of designing and delivering autonomous multi-agent systems at enterprise scale in commercial environments Expert-level understanding of agentic architecture patterns - able to define the strategy, evaluate trade-offs across approaches, and set the technical direction for large teams Deep knowledge of model orchestration, inference cost optimization, and selecting the right cutting-edge models (commercial and open-source) for each use case Experience architecting large-scale vector data systems and embedding pipelines Ability to design highly secure, privacy-preserving AI systems resilient to hallucinations and adversarial inputs Expert-level knowledge of AI-assisted development tools and workflows (e.g., Cursor, Copilot, or similar) - not just personal proficiency, but the ability to define tooling standards, best practices, and productive setups for engineering teams Deep understanding of LLMOps: model versioning, evaluation pipelines, prompt management, cost monitoring, and CI/CD for AI applications - able to define the LLMOps strategy for the team and ensure production-grade operational maturity Strong cloud architecture experience (AWS, Azure, or GCP) Exceptional communication skills - able to lead technical conversations with both engineering teams and C-level stakeholders (Polish and English) BONUS POINTS IF YOU HAVE Experience establishing global AI governance frameworks aligned with regulatory standards (e.g., AI Act) Experience designing observability layers for enterprise agentic architectures Contributions to open-source projects, published research, or recognized thought leadership in the agentic AI space Experience scaling engineering teams and establishing technical standards across multiple projects or geographies Research indicates that some candidates, especially the most diverse ones, may hesitate to apply for positions if they don't meet all requirements. If you believe you possess the necessary skills, even if not meeting every requirement, we wholeheartedly encourage you to