In-person seats unavailable. Participate remotely via the live stream.

Thank you for the overwhelming response. This workshop is housefull! Registrations are closed for in-person attendance.

The workshop will be live streamed. Stream links will be shared with The Fifth Elephant 2025 edition ticket holders and annual members for remote participation on the workshop day.

Workshop overview

Modern AI Agents are no longer toy chatbots; they read and write company data, call third-party APIs, and collaborate with other agents. Yet most tutorials end before you hit real-world issues like passing user context across dozens of tool calls, handing tasks from one agent to another, or diagnosing a silent failure.

Note

• This workshop is of 3 hours duration.
• It is an in-person and hands-on beginner to intermediate level workshop.
• Beginners are welcome, but must be prepared to catch up on:
  • How to make LLM calls
  • Understand the tool calling API for most providers
    This workshop is better suited for participants with beginner level AI/ML experience.
• 25 seats available for in-person participation.
• Live stream will be available for all participants who wish to attend remotely.

🧭 Agenda

Intro (10 min) – Setting up environments and brief introduction.
Motivation (20 min) - Dissecting production agent systems like the GitHub Copilot pair programmer.
Toy Example (30 min) - Let’s build such a system from first principles using either Google ADK or OpenAI Agents SDK.
When multi-agent makes sense (30 min) - Learning why multi-agents make sense in certain scenarios and understanding common patterns.
Break
State Management & Lifecycle hooks (30 min) -Understanding how memory works in agents and how to manage pre/post-tool and pre/post-agent callbacks for validation, logging, and enrichment.
Tracing & observability (20 min) – Emit spans to OpenAI Traces or LangSmith; replay and inspect.
Evaluation loops (15 min) – Define golden tasks; run LLM-as-judge scoring; track deltas over time.
Wrap-up & next steps (5 min) – Checklist for taking prototypes to production; resource list.

💻 Prerequisites - install on your laptops before the workshop

• Python 3.11+ and Docker
• Git & a GitHub account
• Keys for OpenAI (or equivalent)
• Optional free accounts: LangSmith
• Basic comfort with REST & JSON

👥 Who should attend

• Engineers bringing agents into customer-facing products
• Technical founders evaluating multi-agent workflows for automation

👨 🏫 Instructor bio

Jayesh loves exploring new technologies and putting them to work to solve real-world problems. He adapts very quickly to emerging frameworks or languages and plans his days to maintain a balance between sports, literary works, and coding. Jayesh enjoys speaking in public and interacting with people with different experiences than his own.

How to attend this workshop

This is a free workshop, thanks to our sponsors - Composio. Registration is necessary to block seats for in-person attendance.

Seats are limited and available on a first-come-first-serve basis. 🎟️

Contact information ☎️

For inquiries about the workshop, contact +91-7676332020 or write to info@hasgeek.com