A clear guide to prompt and context engineering and an important reminder that they are not the same thing when building AI products. Context isn’t just an input; in many cases it *is* the product: the business logic, differentiation, and core value. PMs should absolutely be part of these conversations too, not just engineering teams. Thanks for sharing this, Karo!
I thoroughly enjoyed your insight here. I think this expands across professional and industry borders. I’ve always hated people selling prompts because it felt like snake oil. I teach people to fish, not give them a fish. I hope you revisit this topic in 6 months and see how things have changed. The idea of the context being built in is exciting.
I especially like the shift from prompts to context as product architecture.
In my own work, I keep arriving at the same problem from the governance side: the issue is not only what the model is asked, but what the interaction is allowed to carry forward, ignore, compress, or stop.
A clear guide to prompt and context engineering and an important reminder that they are not the same thing when building AI products. Context isn’t just an input; in many cases it *is* the product: the business logic, differentiation, and core value. PMs should absolutely be part of these conversations too, not just engineering teams. Thanks for sharing this, Karo!
My pleasure, thank you so much for reading Just! 🤗
Great stuff!!! I loved the side-by-side comparison between the two!
Thank you so much for reading! 🤗
I thoroughly enjoyed your insight here. I think this expands across professional and industry borders. I’ve always hated people selling prompts because it felt like snake oil. I teach people to fish, not give them a fish. I hope you revisit this topic in 6 months and see how things have changed. The idea of the context being built in is exciting.
I completely agree, most of the prompts being sold online are snake oil. I dislike that too. I wrote more about it here, because some of these prompts can actually hurt the output: https://karozieminski.substack.com/p/ai-prompting-techniques-reasoning-models-2026
And thank you for reading Ericka!
My pleasure.
Very solid and comprehensive!
Thank you so much for reading Xian! 🤗
This is so thorough. Thanks Karo! 🙋🏻♀️
Excellent piece, Karo.
I especially like the shift from prompts to context as product architecture.
In my own work, I keep arriving at the same problem from the governance side: the issue is not only what the model is asked, but what the interaction is allowed to carry forward, ignore, compress, or stop.
That is where “prompting” stops being enough.