Description of the talk: - Background. Why we decided to add an assistant to the standard search. - Market Trends: How Agentic Commerce Is Evolving and Why It Matters Now. - Implementation: How we technically built and integrated the AI agent into Prom’s infrastructure. - MVP and experiment results. What the initial tests revealed and how users interacted with the assistant. - Key insights and conclusions. The main takeaways we identified during development and launch. - What's next. Plans for developing AI tools on the marketplace.

In this talk, we will raise several important topics: - Is code really such an important business asset? And if not, why are developers still getting paid. - How exactly Node.js developers can increase their efficiency using "stupid" AI based on GPT-4 to perform their tasks. And we have to go beyond "Write the documentation on this code". - Who are the Prompt Engineers, and why can you be fired in a year because of them?