Free Webinar
Building Agents with Granite Workshop
In this workshop, we will explore writing AI agents using IBM Granite AI models. We will touch on what are AI agents and explore some different designs for agent development. These designs are informed by the experiences of IBMers in developing agents. Our first agent will be a basic Function Calling Agent where we will build the agent from scratch to learn how agents work. We will then explore other agent designs such as Plan-and-Solve, Route-and-Solve, ToolRAG, and ReAct. The workshop agents will use LangGraph with LangChain in their implementations and use IBM Granite as the language model.
What You Will Learn
- Introduction to AI agents and agent design patterns
- Building a Function Calling Agent from scratch
- Plan-and-Solve, Route-and-Solve, ToolRAG, and ReAct agent designs
- Hands-on with LangGraph, LangChain, and IBM Granite models
- Real-world agent development experiences from IBM
Speakers
Mark Hinkle
Host, All Things AI
BJ Hargrave
Senior Technical Staff Member, IBM Research
Register Free
Opens WebinarJam registration. Takes 30 seconds.
Or register below
More Upcoming Webinars
Continue learning with these free live sessions from the AIE Network.
The Path to Becoming AI-First: An Operator's Playbook
Tuesday, March 10, 2026 at 12:00 PM ET
What happens when a recruiting operator — not a software engineer — decides to rebuild his company around AI? In this live lunch & learn session, James Horne, Director of Delivery at Axiom Path, shares how his team built 25+ AI automations and digital employees that now operate alongside human recruiters — screening candidates, managing workflows, and turning 21,000+ historical recruiting notes into a searchable knowledge system. James will demonstrate a live voice AI built to talk about itself and its role in Axiom Path and their AI Transformation. This session explores what it really means to become AI-first: starting with process, increasing leverage, and learning how to manage digital employees alongside human employees. This is not a coding talk. It’s an operator’s blueprint.
View detailsSecuring Your Model Context Protocol Footprint
Tuesday, March 17, 2026 at 12:00 PM ET
Leaders across industries are using the Model Context Protocol (MCP) to put tools in the hands of their AI agents, so they can do real work and drive measurable return on AI investments. However, it’s critical that MCP be implemented with guardrails to protect against malicious actors and unintended actions. Stacklok will show you how to use its open source ToolHive MCP platform to curate a trusted MCP registry and implement auth, token exchanges and other security best practices so you can put MCP into production.
View detailsOps & DevSecOps Guardrails for Agents
Wednesday, March 18, 2026 at 12:00 PM ET
Agents writing code is one thing, but agents involved with infrastructure are something else entirely. In this final session of our lunch-and-learn series, we move from developer workflows into operations — like CI/CD pipelines, cloud environments, Kubernetes clusters, observability systems — and the governance questions that follow. What does safe autonomy look like when an agent can read signals, form a hypothesis, and propose (or execute) change? This session is about guardrails before scale. We will explore how ops-focused agents actually behave in modern environments, the risk in places teams are already exposed, and baseline controls that make experimentation possible without increasing blast radius.
View details