When we give gen-AI a task, we’re told to give it context. Context is king! Context will make Claude work for you. Context will solve all the world’s problems. But the context we’re giving it is actually making things worse.

This often happens because the context fed into the machine is bad. Too much, too broad, or formatted in the wrong way - it creates confusion and leads to unexpected, inconsistent, or generic output.

In a context design method I’ve developed, I found that putting together context in a structured, well-designed way is the first and most important step to getting value from AI. Structured context gives models the right information to work with and reduces the noise that leads to generic results.

Through stories and before-and-after examples from my own work and hands-on excercises, I will show why AI so often misses the nuances that experienced professionals resolve in seconds. I will introduce Context Design as an emerging practice that shifts us from writing one-off prompts to building an intelligent information infrastructure grounded in the professional knowledge people already hold.

Participants will learn how to build and maintain reusable context design packages with a strong ROI, that support more consistent, high-quality work across teams.

First done as a talk at the UXI conference in 2026, this session received excellent feedback. It was then done again in a shortened capacity at The 3rd Loop AI event in Tel Aviv, again receiving great reviews. Talk video available (in Hebrew with English subtitle): https://youtu.be/mKx6hXgo8e4

Some of the feedback received:

"Excellent! I was fascinated"
"Clear, practical, and fun"
"Excellent topic! And very valuable today with increasing AI use"
"Loved how practical it is"