Four months ago, I proudly shared scale testing results for an early private preview version of GraphnAI. ~1M nodes and ~1.8M edges ingested and analyzed in just under 17 minutes!
GraphnAI has evolved significantly since then, with 3 additional patents filed covering analysis that goes much deeper. The edge count grew 3x, and that's AFTER our patent-pending edge consolidation cut the total number of edges by ~50%.
The focus throughout was accuracy. Analysis stayed fast at small scale, but we knew we needed another round of large-scale testing. Some increase was expected. The initial results were not: analysis at the ~1M node scale had grown over 10x, taking ~177 minutes to complete.
I refused to accept that, and the refusal paid off. We've brought analysis down to just over 19 minutes at the ~1M node and ~6M edge scale. That's analysis far more complex and complete, with more features and capabilities than four months ago, at the cost of only ~2 extra minutes.
In that post four months ago, I may have said that unlike Domino's๐in the 80s, I wouldn't be making a "30 minutes or less" guarantee... but I'm happy to see we can still deliver time-to-value of 30 minutes or less (often far less!) even at massive scale.
#BuildingInPublic #IdentitySecurity
With all the VC horror stories I've read on here lately, I'm glad my VC experiences have been mostly positive. Then again, I haven't started actively raising yet.
Since announcing the founding of GraphnAI at the beginning of the year, I've met with a half dozen or so VC firms, a few bigger names, a few smaller. All interested in investing in a Series A, one interested in Series D. All of them contacted me first and said they like to establish relationships a year or more in advance. Most of them virtual meetings, but a couple in-person too. The VCs I met with were all very attentive and seemed genuinely interested.
Because GraphnAI is currently pre-revenue and more on the line between Pre-Seed and Seed, and all the VCs I've met with so far only do Series A or later, I did apply to both YC and a16z speedrun. Rejected by both with the same "we encourage you to try again later" form e-mail most people applying to them get.
The new Nvidia DGX Station for Windows
The Ultra Desktop comes with up to 784GB of coherent memory and up to 20 petaflops FP4 compute
It can handle up to 1 trillion parameters locally
Systems based on it are available priced at up to around $85,000
@SwiftOnSecurity@_numinit I've had Clade tell me it found a typo in something it wrote itself just a few seconds earlier... wild that it hallucinates typos, but not super surprising when you consider their training data is likely full of them.
Every SWE I've ever worked with (outside of the ones I worked with at Microsoft) hated developing on/for Windows, to the point of actively refusing to do it unless they absolutely had to. So, it's not really surprising that LLMs suck at it when they were trained on/by humans with that kind of attitude.
@_rygo6 Been there for a while. Building from a solid foundation of domain expertise and running experiments to prove/disprove theories is extremely powerful.
@IceSolst I've been meeting with multiple NYC VC firms regularly since January. Not one of them ever mentioned "NY Tech Week", but they did ask if I would be at RSA a few months ago. I only discovered the existence of "NY Tech Week" from a post on here last week. ๐คท
@DarthMaulware@ImposeCost Illinois is trying to pass the same thing right now. This state has a habit of outlawing everything I own. Luckily they've included a grandfather clause, so I remain legal. Just can't buy new stuff. ๐
They botched the rollout as always. Completely unusable. Tried forcing back to 4.7, got a little further, but still running into the same API Error. Seems lots of people are experiencing the same issue right now, and yet their status page says everything's great. https://t.co/agZ8rh30hs
Working on stuff in Claude Code without issue. Start a new session and "SURPRISE! Opus 4.8!". Cool, except every prompt results in this API Error. Completely unusable.
Keep in mind these types of "requirements" are often just wish lists. Unfortunately, the people (or nowadays AI) performing the initial screening often don't understand that experience with a similar class of application and deep understanding of the fundamentals behinds it means you'll quickly and easily pickup the specific application they use.
Don't eliminate yourself, but don't hold your breath either.