Course Objective & Scope
As different teams across an enterprise build specialist agents on LangGraph, OpenAI Agents SDK, CrewAI, Copilot Studio or Vertex AI, they need a standard way for opaque agents to discover one another and collaborate without sharing internal state or prompts. This course covers the open A2A (Agent2Agent) protocol and enterprise agent interoperability architectures.
Labs cover the architectural boundary between MCP (agent-to-tool) and A2A (agent-to-agent), publishing agent.json Agent Cards, the stateful Task lifecycle, streaming updates via SSE, webhook push notifications and OAuth/mTLS security. Participants connect two agents built on different frameworks into a single A2A workflow.
Who Should Attend?
- AI architects designing organisation-wide multi-agent ecosystems and platforms
- Software and platform engineers integrating agents across teams or vendors
- Integration specialists managing enterprise API gateways and service meshes
- Prerequisite: REST/JSON-RPC, async messaging and experience with at least one agent framework
Key Takeaways
- Decide architecturally when to use MCP (tool access) versus A2A (peer agent delegation)
- Author standard Agent Cards declaring skills, modalities and authentication requirements
- Implement the stateful A2A Task lifecycle including multi-turn clarification (input-required)
- Handle long-running cross-agent tasks asynchronously with SSE streaming and webhooks
- Secure inter-agent communication with OAuth 2.0, on-behalf-of (OBO) token exchange and AI gateways
Curriculum
Interoperability in a Multi-Framework Enterprise
- The challenge of collaborating across opaque agents without shared memory or prompts
- How MCP and A2A complement each other: the tool layer versus the agent layer
The A2A Protocol: Agent Cards and Discovery
- Structure of
.well-known/agent.json, skill declarations and input/output modes - Internal Agent Registries and dynamic capability discovery
Task Lifecycle, Streaming and Asynchronous Push
- Task, Message, Part and Artifact data structures
- State transitions: working, input-required, completed, failed and canceled
- Live progress via SSE streaming and webhook push notifications for long tasks
Security, Identity Delegation and Cross-Agent Tracing
- Inter-agent OAuth 2.0, mTLS and preserving user permissions via On-Behalf-Of (OBO) tokens
- Propagating W3C Trace Context and enforcing budgets across agent boundaries
Workshop: Bridging Two Frameworks with A2A
- Connecting a LangGraph agent with an agent built on another SDK over A2A
- Testing multi-turn clarification (input-required) and failure handling
Other Courses in This Category
-
Training on pausing agents for human approval on high-risk tool calls and low-confidence decisions, editing tool arguments, routing asynchronous escalations via Slack/Teams/web UIs and maintaining audit trails.
- Live Course
- 5 modules
-
Hands-on training for developers on building ReAct agents from scratch: the Thought–Action–Observation loop, native tool calling, stop conditions, iteration limits and failure analysis.
- Live Course
- 5 modules
-
Technical course on agents that split multi-step tasks into planner, executor and replanner roles, covering ReWOO and LLMCompiler-style parallel task graphs, plan validation and cost optimisation.
- Live Course
- 5 modules
-
Course for teams building long-running agents: state schemas, reducers, checkpoints, thread management and state-machine workflows, illustrated with LangGraph and workflow engine examples.
- Live Course
- 5 modules
-
Hands-on course on designing tools agents can use reliably: naming, descriptions, JSON Schema parameters, error messages, tool contracts, versioning and permission scopes.
- Live Course
- 5 modules
-
Hands-on technical course on building production-ready agent systems with LangGraph: StateGraph, conditional edges, checkpointers, interrupt-based human-in-the-loop, subgraphs and the Send API.
- Live Course
- 6 modules
Get in touch
