Berlin AI Engineering Group with m|rig

******************************************************

Please do not forget to register, or you might not be able to join the meetup next week. Here is the Teams link to register for the webinar:

https://events.teams.microsoft.com/event/7bb572ef-b99e-40a0-a039-ec01c7620f7a@d94ea0cb-fd25-43ad-bf69-8d9e42e4d175

******************************************************

Planning a trip can quickly become complex:

• Comparing flights across platforms
• Selecting hotels based on budget and preferences
• Organizing activities and optimizing schedules

While many tools provide partial answers, few are able to combine all constraints into a coherent, optimized travel plan.
This is where Agentic AI systems unlock a new level of automation.
But building a reliable AI travel planner requires more than just querying APIs with an LLM.
How do you design a system that can:

• Retrieve and combine data from multiple sources (flights, hotels, activities),
• Reason over constraints like budget, time, and preferences,
• Plan multi-step itineraries,
• And coordinate multiple agents working together?

In this 60-minute interactive webinar, we’ll walk through the end-to-end design and implementation of an AI-powered travel planner, using advanced agentic design patterns and multi-agent collaboration.
The session will include a live implementation using Python, LangChain, LangGraph, OpenAI, and Google APIs, demonstrating how to orchestrate planning agents, tools, and workflows to generate a fully optimized travel itinerary.

***

### You’ll learn

✈️ Core Concepts of Agentic AI for Planning Systems
How planning problems differ from simple Q&A:

• Multi-step reasoning and constraint satisfaction
• Tool usage and API orchestration
• State and memory across complex workflows
• Deterministic vs agent-driven planning

***

🧠 Advanced Agentic Design Patterns
Explore key building blocks of modern AI systems:

• Tool-based agents (APIs for flights, hotels, maps)
• Planner–executor architectures
• Multi-agent collaboration (specialized agents per task)
• Human-in-the-loop validation for critical decisions
• Advanced prompt engineering for structured planning

***

🔗 Multi-Agent Orchestration with LangGraph
How to structure complex workflows:

• Defining agent roles (planner, search, optimizer, validator)
• Managing state transitions across steps
• Handling failures and retries
• Ensuring consistency across agents

***

🌍 Live Demo: Building an AI Travel Planner
Step-by-step walkthrough:

• Connecting to external APIs (Google services, travel data)
• Designing a planner agent to structure itineraries
• Coordinating sub-agents for flights, hotels, and activities
• Optimizing budget allocation across the trip
• Generating a full, structured travel plan

***

💡 From Prototype to Real Product
Discussion around:

• Reliability and hallucination control
• Cost optimization (API + LLM usage)
• UX considerations for travel assistants
• Scaling to production systems

***

📅 Duration: 60 minutes
🛠️ Tech Stack: Python, LangChain, LangGraph, OpenAI, Google APIs