I warmly invite you to watch this short clip - it so positively articulates what we as #devrevelers have been working on - it's amazing to be part of this journey.

Computer WORKS hanks to deep integrations (#airsync) with existing structured AND unstructured data (computer #memory) and sophisticated orchestration and analytics functions that underpin productive #ai.

Take a look at the new @devrev blogpost that I just published: devrev.ai/blog/smarter... ... as always - I would love to hear your thoughts!

Do I buy something off the shelf?
Or develop everything from scratch?
Or, maybe, just maybe, could it be that there is a middle ground - and that you can buy a platform upon which you can build FASTER?

This article also reminded me of an old distributed systems discussion that I learned in my uni days - how distributed "sync"(chronous) systems are so hard to get working at scale, and how "async" systems are conceptually harder, but much more resilient.

Join us by registering over here lu.ma/x2n8bn33 - it's will be a great event.

So how do you make that tangible? How to you explain that? Well, as usual, *show*, don't tell, is the better method. And that's what Sunil and Richard are going to be doing during our webinar next week.

What's the best approach? In my view, both #build and #buy are dead: there's a happy medium between the two approaches, where you buy the core stack, and then customize that in detail. Hope you like the article. Ping me if you want to discuss.

That, I promise, is not a small task. And that, I offer, is one of the main value drivers of the @hopsworks.bsky.social #AI #Lakehouse for our customers.

There's a lot of great "pieces of the puzzle" out there that you can just pick off the shelf, install in your environment, and get going. At least that's what it looks like, because you still need to *integrate* the different pieces, and then *keep them integrated* forever thereafter.