KEY LEARNING OUTCOMES
By the end of the programme, you will be able to architect, develop, evaluate, and deploy robust, deterministic multi-agent systems that operate reliably at enterprise scale.
-
Understand the evolution of AI: Trace the progression from predictive and generative AI to frontier models and enterprise agentic systems.
-
Apply enterprise knowledge methodologies: Implement the OKF methodology to assess, structure, and transform enterprise information into AI-ready knowledge.
-
Assess AI-enabled data readiness: Use AI-assisted techniques to improve data quality, metadata, governance, and access controls for reliable execution.
-
Build grounded knowledge bases: Design stateful, hybrid RAG and GraphRAG architectures connected to trusted enterprise data sources.
FOUNDATIONS, KNOWLEDGE ARCHITECTURE & DATA READINESS
-
Design agent harnesses & loops: Define execution environments, tool schemas, permissions, and plan-act-observe-evaluate cycles.
-
Build deep agents & long-running workflows: Combine planning, memory, context management, and delegation for complex multi-step execution.
-
Orchestrate multi-agent architectures: Use LangGraph to build multi-agent systems with clear routing, delegation, and verification patterns.
-
Structure tools & schemas: Validate typed inputs and outputs, build modular MCP servers, and integrate securely with enterprise APIs and databases.
-
Monitor and evaluate agentic systems: Trace decisions, tool calls, and state transitions using LangSmith to ensure auditable and deterministic performance.
MULTI-AGENT SYSTEMS, LOOP & HARNESS ENGINEERING
-
Deploy durable production systems: Implement persistent checkpointing, state recovery, idempotency, and operational resilience to handle system interruptions.
-
Master model routing & cost management: Track token usage, enforce budget controls, and dynamically route requests across frontier models based on latency, cost, and capability.
-
Implement safety & human oversight: Apply guardrails, permission boundaries, and Human-in-the-Loop review before executing consequential actions.
-
Navigate enterprise governance & ethics: Translate regulatory frameworks like the EU AI Act into practical monitoring, auditability, and responsible autonomy controls.