@1littlecoder @nebiustf @NVIDIAAI @tavilyai @LangChain Nice one!

As you can see from the demo, Signals of AI deepagent sourced 45 sites to prepare the dossier.

This is just the beginning, it's an ideal skeleton for anyone to make a product out of it!

https://github.com/amrrs/signals-of-ai-deepagent

@1littlecoder @NVIDIAAI @tavilyai @LangChain Great work 🙌

@1littlecoder @nebiustf @NVIDIAAI @tavilyai @LangChain How are you ensuring the data that is returned is real? Website started blocking the crawls or feeding the garbage instead of real data. Cloudflare made it even more easier these days.

@1littlecoder @nebiustf @NVIDIAAI @tavilyai @LangChain very interesting project! thanks for sharing

@demian_ai @nebiustf @NVIDIAAI @tavilyai @LangChain thank you !

@nebiustf @NVIDIAAI @tavilyai @LangChain thank you ☺️

This is the layer I am most curious about with competitor-analysis agents. On the SME deployments I have run, the wall is signal over coverage: a deepagent will happily surface forty competitor moves a week and the team ignores all forty. What made it usable was a relevance filter tuned to the two or three strategic questions the team actually acts on, so it reports the three that matter rather than everything it can find.

@HowDevelop @nebiustf @NVIDIAAI @tavilyai @LangChain thank you for checking it out :)

@aka_yaswanth @nebiustf @NVIDIAAI @tavilyai @LangChain The report is grounded on the search data, but I think I can probably look into the validation