HOT STARTUPS AI search infrastructure

strap in, there is a huge and explosive market building search infrastructure for AI agents. So, this is the new world we're living in and let's dig into three different companies which have raised a ton of money and are doing very interesting things. Let's start with access. They recently raised 250 million at a 2.2 billion dollar valuation led by Andris and Horwitz. Of course, they actually started five years ago, team out of Harvard. It was before GBT even existed. The founders originally wanted to just build a better search for people and then with the release of GBT AI companies started reaching out for API access and they pivoted fully into AI infrastructure which has been massive for them. Today they have over 400,000 developers 5,000 companies using XAs APIs including cursor, hotspot and open router which also just announced a huge raise yesterday. They've also built their own neural search database and deployed uh 144 H200 GPU so agents can search the web uh in under half a second which is crazy. One AI company actually reported told XAs that if cost and latency were solved their agents basically search 100% the time. So, uh let's dive into the next one which is parallel that you've probably heard of. They raised $100 million dollars recently at a 2 billion dollar valuation like by. I was actually founded by an ex-Twitter CEO Prog after Eon acquisition Twitter in 2022. Instead of focusing on search speed, parallel was building a deep research API that helps AI agents do complex multi-step research tasks. So, these are obviously different use cases all related to search, parallel being more complex. Um and ex being more basic. Super interesting. One of our customers is actually Harvey the legal AI start up. And instead of a lawyer manually having a date through hundreds of documents for hours and AI agent can actually research like the case lab, summarize their findings and automatically trace the sources. Last but not least, and I will say that I'm an investor in this one, I co-led the seed round um earlier this year. is a company called 9 N. It's a newer startup in the space. They raised a $5.3 million seed round, they were part of South Park Commons. They were approached a problem from a totally different angle. So it was founded by Michael and Mod. they're a father son duo. and they had the idea that AI agents still don't really understand humans and consumer profiles. So, they're building what's called a people graph for AI agents. They are software connects information across platforms like LinkedIn, Instagram, X, Spotify, Strava, they want to know everything about you to help them understand your interests, your behavior, um and how you're interacting online. So, this is interesting because it's continuous. A lot of data that you can get from like let's say AI info, which by the way, could partner with them. Um is very very dated, not always up so it needs to be refreshed constantly so you feel like you're getting the latest. Uh nine can have a monthly kind of dump and updates that we uh know more real-time information on each person. Um an example of this would be instead of an AI uh assistant only knowing your calendar and emails, it could understand that you like to listen to house music when you're working, you like to run marathon, do hiroxes, uh you follow startup investing and then Taylor the response to you with that context. Companies like XSA and Parallele are building infrastructure that helps agents retrieve and reason over information at scale and nine just gives you context on the humans that those agents are acting on behalf of. So very interesting times. I think these AI for plays are very fascinating. They've raised tons of money and I just be on the look out for what's to come next year.