​Building Better Agents

​From Hype to Helpful

​Join this hands-on workshop exploring what it really takes to build AI agents that people can trust.

​Most agent demos look magical for five minutes, then feel chaotic in practice. In this session, we'll break down the engineering patterns and design principles that separate impressive demos from production-ready systems.

​What We'll Cover

​Understanding Agent Architecture

• ​What makes something an "agent" vs. a chatbot or script

• ​The decision-action loop and why it matters

• ​Common failure modes and how to avoid them

​The Four Pillars of Reliable Agents

• ​Clear intent: defining boundaries and constraints

• ​Good memory: what to remember vs. what to forget

• ​Disciplined tool use: when not to act

• ​Feedback loops: correction and improvement over time

​Autonomy vs. Trust

• ​Why predictability beats flashiness

• ​Human-in-the-loop as a feature, not a bug

• ​Earning autonomy incrementally

​Practical Patterns

• ​Designing failure states first

• ​Shipping observability before features

• ​Building narrow, focused agents that work

​Who Should Attend

• ​Developers building with LLMs or agent frameworks

• ​Product managers designing AI-powered features

• ​Founders exploring agent-based products

• ​Anyone curious about making AI feel less chaotic

​No crypto knowledge required — we'll focus on universal agent design principles applicable across any stack.

​What You'll Leave With

• ​A clear mental model for agent architecture

• ​Concrete patterns for trust-first design

• ​Practical examples and anti-patterns

• ​Resources and code samples to explore further

​About the Hostess

​Val Alexander is Lead DevRel Engineer at Ritual, where she works with teams building AI-native applications. She spend their days seeing what works, what breaks, and translating builder feedback into better tools, docs, and patterns.

​Before Ritual, she was a Senior DevRel engineer at Morpho (2nd largest DeFi lender) and Chainlink Labs (largest oracle provider). She also cofounded a decentralized exchange in 2021 and brings almost a decade of professional experience in blockchain and machine learning engineering experience.