yggy
The name is odd, the pitch is one line, and yet I'm genuinely curious about this one. An AI knowledge graph that builds itself is the kind of thing that sounds like a developer's fever dream until you realize how much time actually goes into manually mapping entities and relationships. If yggy can turn that into a prompt-and-go experience, it solves a pain point that every data team I've talked to complains about.
What I'd want to know is how it handles the messy edge cases. Knowledge graphs are easy to demo with clean data, but the real world is full of ambiguity, duplicates, and shifting schemas. Does it learn from corrections, or does it just guess and move on? The difference between a toy and a tool lives in that feedback loop.
I think the bigger story here is the timing. As more companies bolt AI onto their products, they need structured context to make those models useful. yggy is betting that the graph can assemble itself faster than a human can. That's a bet I'd take a second look at, but I'd want to see it run on genuinely messy data before I get too excited.