Paul Querna

I believe: 1) Corps will use 100s of different SaaS products (10x-100x from today) 2) New Arch required to enable #1: A fabric between companies, humans, and SaaS 3) Identity is going to become about Automation, Orchestration and Delegation -- the fabric between the parties

It was all fun and games until the SQLite VACCUM took 12 hours

Claude: THE PATH IS CLEAR Me: we are absolutely not making it out of this alive

Calling it now, PowerShell is more token efficient than bash, and someone else will figure this out in the next 3 weeks.

Screw Miami, lets all move to Hawaii.

If your leadership team isn’t coding you’re NGMI

"we could just fork postgres" escalates really quickly when everyone has AI-agents working for them

2 days ago prediction: that's going to be an 80,000 line PR current status: +78,840 −890

“We can’t work with that startup because they are not SOC 2 Type 2 compliant” “Hey Moltbot, here is my credit card info and all my passwords. Do something cool. Make no mistakes”

0% chance you can explain the state of AI to anyone outside of this website and not look like this right now

the greatest gift to software engineering in the last 3 minutes was the realization you can just make claude deal with bazel.

@AstasiaMyers The greatest conundrum Should your docs be public so AI learns your product, or should they be private for the marginal learnings edge

AI makes code cheap, but product learning isn’t. The real bottleneck is knowing what to build next. This can come from customer adoption and feedback. Product Manager AI agents are the next phase to helping synthesize customer signals, adoption, and feedback into better inputs for building.

@AstasiaMyers But I might vibe code my own CMS

The ‘everyone will vibe-code their own SaaS’ take misses why SaaS won in the first place: deep understanding of workflows, systems of record, team coordination, edge cases, and maintenance. AI lowers build costs. It doesn’t erase product expertise.

I’m not addicted to my phone, I’m addicted to Agents

if you are vibe implementing a client, and someone else is vibe implementing a server, and they both hallucinate the same JSON field, maybe we should just change the RFC.

Every senior engineer I know is currently bottlenecked by code review, correctness, and trying to wrangle the output of their team in sound architectural directions

are we still in the slow takeoff, i can't tell

Today, ProsperOps announced that they are being acquired by @flexera . This is an incredible outcome for ProsperOps and a seminal moment for us at Active Capital. ProsperOps has been a special company for me and our firm from the very beginning. I worked with the founders for many years at @Rackspace. We weren’t especially close at the time, but I had deep respect for their work. In 2018, after two of the founders left Rackspace, we met quietly in New Braunfels to talk through the idea of starting something new. A few months later I led their pre-seed round at Active Capital. I loved their vision. From my time in the cloud world, I knew how massive cloud spend was becoming and how painful cost optimization was for businesses. Most tools at the time focused on reporting. ProsperOps had a very different idea: automate savings and optimization directly. Even before AI was fashionable, the company was built around automation, intelligence, and outcomes rather than dashboards. From there, the company took off. What stands out in hindsight is how intentionally ProsperOps was built. Under their leadership, this became one of the best-run companies I’ve been involved with. The culture was real and durable. You could see it in employee tenure, public sentiment, and consistent execution over time. These guys built the company from first principles, and I have incredible respect for that. For example, the company started generating revenue quickly, and as growth accelerated, there was pressure to raise more capital. They resisted that. It simply wasn’t necessary. In fact, ProsperOps never raised another venture round beyond our initial investment yet was always one of the fastest growing companies in our portfolio. Then came early acquisition interest. When you’re only a couple of years in and someone offers what feels like life-changing money, it’s hard to turn down. There was understandable temptation. But instead of selling early, we provided secondary capital so the founders could take some liquidity off the table and keep building. I’m not sure how many pre-seed venture capital firms would do that. But it changed everything. By staying independent, ProsperOps continued to compound. Over time, the company became far more meaningful and far more valuable. This is the biggest outcome for our firm to date. I feel incredibly proud and privileged to have been the lead investor alongside this team from the very beginning. Watching a group of former Rackers go on to build one of the most successful companies to come out of our ecosystem is something I’ll always be grateful for. I want to specifically thank the three founders: - Chris Cochran, an incredible leader and CEO. - Erik Carlin, a product savant who could see around every corner. cc @ecarlin - Chris Kuehl, whose technical vision was remarkable to watch play out at scale. Congratulations to the ProsperOps team and to Flexera. And thank you to everyone involved in the ProsperOps journey.

Aesthetic: Dark-mode, cyberpunk, trains, hexagons, puppies with jobs AI does know me.

I’m ready to be an executive with an iPad Ajr.

Unbelievable to me how many people still don’t understand this.