Rachel Warren

So you've kind of got the bear case, which is that AI might eventually just allow teams to build their own custom workflows.

So that makes a platform like this one redundant.

Now, monday.com and their management are obviously betting that AI will be the glue for their platform.

You know, that new AI work platform I mentioned, it's designed to automate kind of the tedious parts of project management, try to make the software stickier for their clients.

For me, the biggest drawback for this business or the stock, I should say, you know, they're not profitable on a gap basis.

They're still really leaning heavily into stock based compensation.

I think the stock is probably going to be under pressure for a while.

Yeah, so essentially Cerebrus is looking to build a massive alternative to NVIDIA's empire by rethinking what a chip even looks like.

So instead of making thousands of small chips from one silicon wafer, they actually use the entire wafer to create one giant processor, the wafer scale engine.

They say this can train and run AI models far faster than traditional GPUs.

So they...

don't just sell silicon, they package it into these multi-million dollar supercomputer systems that handle everything from power to liquid cooling in a single unit.

And so they make money through a few different channels.

One is selling these massive hardware appliances.

They provide recurring software and support services.

They also offer AI as a service where companies can rent time on their chips via the cloud.

Now they've recently landed very significant deals with hyperscalers like AWS in March.

They signed a binding term sheet with AWS.

That was the first major hyperscaler to commit to deploying Cerebrus systems in their data centers.

They just signed a big deal with OpenAI.