How to Build Reliable AI Agents
From Simple LLM Interactions to Production-Ready AI Workflows
Davy helps developers build high-performance, maintainable applications. He has over a decade of experience as a software engineer, working with teams across the world to design and deliver modern software systems. For the past two years, he has been focused on AI-native applications, developing practical best practices for building reliable, production-ready AI agents.
About This Workshop
AI agents are easy to demo, but much harder to make reliable in real-world applications. In this hands-on workshop, Davy will walk participants through the core building blocks behind modern AI agent systems and explain how to design them in a more robust, predictable, and production-minded way.
The session will cover how agents use tools, how agent loops work, how to manage context effectively, and when to introduce human-in-the-loop approvals. Participants will also explore the practical challenges of reliability, evaluation, and system design as they move from simple LLM interactions to more capable AI workflows.
This workshop is designed for developers seeking a practical introduction to building AI agents that do more than just look impressive in a demo.
What You'll Cover
Eight key topics covering the full spectrum of building reliable AI agents.
What AI Agents Are
Understand what distinguishes AI agents from simple LLM applications and why that distinction matters.
Tool Calling
Learn how to connect agents to external systems and APIs through tool calling patterns.
Agent Loops
Understand how agent loops and task execution flows drive autonomous behaviour.
Context Management
Explore context and memory management strategies that keep agents on track across long tasks.
Human-in-the-Loop
Design safer systems using human-in-the-loop approval patterns at the right decision points.
Reliability Challenges
Identify and address the common reliability pitfalls when moving agents from demo to production.
Evaluation & Improvement
Learn how to evaluate agent behaviour and iterate toward more predictable, robust systems.
Production Design Patterns
Practical patterns for designing production-minded AI agents that do more than just look impressive in a demo.
Session Agenda
A focused 1.5-hour session moving from fundamentals to production patterns.
Introduction to AI Agents
15 minutesEstablish a shared understanding of what AI agents are and how they differ from simple LLM applications.
Tool Calling & External Systems
20 minutesHands-on look at how agents call tools and connect to external APIs and data sources.
Agent Loops & Context Management
20 minutesExplore how agent loops drive task execution and how to manage context effectively across turns.
Human-in-the-Loop & Reliability
20 minutesDesign patterns for safer agents and practical strategies for improving reliability in production.
Evaluation & Production Patterns
15 minutesRound up with evaluation techniques and design patterns for taking agents to production.
Who Should Attend
- Software developers experimenting with LLMs
- Technical leads looking to move toward agentic workflows
- Builders who want to understand how modern AI agents work
- Developers seeking practical, production-minded AI design patterns
Prerequisites
- Comfortable with general software development concepts
- Some familiarity with APIs and modern application development
- Prior experience with LLMs or AI tooling is helpful but not required
What You'll Take Away
- Understand the key building blocks of AI agents
- Know how tool calling, loops, and context handling fit together
- Recognise the main challenges involved in building reliable agents
- Learn practical patterns for designing safer, more robust AI workflows
- Leave with a clearer mental model for building production-ready AI agents
Reserve Your Seat
20 spots available · April 21, 2026 · 17:00–18:30 · Smallpdf AG, Zürich
- 1.5 hours of hands-on training
- Workshop materials included
- Practical design patterns for production AI agents
- Networking with developers & builders
Great! Now choose your payment method:
Secure payment options. All prices include VAT.
Frequently Asked Questions
Ready to build agents that actually work in production?
Join Davy on April 21st and leave with a clear mental model for building reliable, production-ready AI agents.