Serroean retweeted
Excited to share xApple and @a16z are teaming up to host “Entirely New,” an evening bringing together ex-Apple founders, leaders, and superallies, on Oct 19th.
The evening will feature a fireside chat led by @joshelman and @mabb0tt, with category-creating founders leading some of the hottest startups in today’s rapidly changing landscape.
Bryce Schmidtchen from Reactor
Sean McCarthy from BackOps
Davide Asnaghi from Diode Computers, Inc.
and one more panel!
Each of these startups are part of the xApple community, backed by a16z, or both:
Thank you Josh Elman and michael abbott for making this collaboration happen for our community. And a shoutout to Han Nguyen @davidbooth Marci Rosalez at A16Z 🙏
#xapple #xapplefounders #a16z #startup
Some of the best engineers, designers, product thinkers, and founders came out of Apple. In partnership with the xApple community, our team at @a16z is getting them together all in one room.
On Oct 19th in SF, I'm moderating a panel of founders who are building something entirely new across infra, software, and more.
xApple alumni, come hang 📷
partiful.com/e/MIHEcHKbjXTcm…
Serroean retweeted
GPT-Live-1 is now available in the API.
Bring ChatGPT’s natural back-and-forth to your app, with voice agents that listen while they speak and work with the models and harness you choose.
Serroean retweeted
"people think TPM is gonna be the job that gets replaced first. in reality, it's gonna be the hardest thing to replace."
new Dr. Niklas episode. I talk to @jakesrro (founder of Serro AI, ex-Apple engineering manager) about why every vertical agent makes the horizontal coordinator MORE essential, the client with 2 TPMs and 75 recurring reports, and the temporal knowledge graphs that give agents organizational memory.
DN #31: Agentic TPM, Temporal Knowledge Graphs & The Job AI Can't Replace (w/ Jake Kim)
Subscribe to join the build. 🚀☘️
🎧 Listen to the full episode here:
📺 YouTube: youtu.be/sRuQNAyiJag
🍎 Apple: podcasts.apple.com/us/podcas…
🟢 Spotify: open.spotify.com/episode/3P3…
Serroean retweeted
We came a long way as a company. After 2 years of pivot-hell, our message is finally consistent. Not a message we think everyone will get immediately, but something we believe in fully. Believing in it fully means, we also understand it’ll take some time for others to start sharing our vision.
Agentic TPM is going to be the infrastructure that unlocks the next level of AI-native engineering. Enabling real-time collaboration, many-humans-to-many-agents.
Serro is building long-running multiplayer TPM agents as collaborative program management infrastructure.
We’re launching in two weeks on product hunt and @nicklaunches with startup pricing. Please keep an eye out for us as we take on the giants and the mega-funded startups!
Serroean retweeted
Jacob Bank, former Google product lead:
"I built up this team of 40 AI marketing agents to work with me. I'm the only marketing person."
In a 15-minute talk, he shows what one person with the right setup now runs alone.
Forty agents. One human. His AI bill is $500 a month, against the $50,000 a human team would cost.
That's the math quietly minting the first solo fortunes of the AI era.
Watch the talk, then read the article to build your first AI team.
Serroean retweeted
Alex Karp went on CNBC and spoke the truth.
Companies need to own their own memory layer or the large model providers will run you out of your own business.
Serro builds the most granular program memory for companies and makes it available for all downstream agents in real-time.
With Serro, your company owns its own alpha.
loom.com/share/46959ff3a5fa4…
Serroean retweeted
Serro is #1 of 76 launches on NickLauches
We're building Agentic TPM.
It's a collaborative, voice-programmable AI that tracks and surfaces all your teams' initiatives. The shared program memory can then be made accessible to downstream workflows, agents, and loops!
Special 85% off offer on Nick Launches for yet-to-be-funded startups because we're investing in you and the startup ecosystem.
@nicklaunches
RT @jakesrro: We launched Serro on @nicklaunches today!
We're building Agentic TPM. It takes live signals from all your tools and categori…
Programs. You hear about them. If you work at a large company, you may have even been part of one.
But the meaning of “program” is changing.
As an industry we've seen the evolution from feature engineering -> project engineering -> product engineering.
Now, as agents accelerate execution, the next operating model is emerging: program engineering.
When execution gets faster, coordination becomes the bottleneck. Teams, systems, decisions, dependencies, and agents all need to stay aligned around the same outcome.
That is what makes programs so important. They are becoming the new unit of execution for a world where humans and agents build together at much higher velocity.
So we think it's time to revisit what a program means, and how it becomes relevant as we move from assigning AI agents tasks to sharing long-horizon cross-functional goals.
#loops #programengineering
Read more here: serro.ai/resources/what-is-a…
Serroean retweeted
Serro blog is live!
The first post is about the big conviction we have as a company. About why every engineering org is going to have hyperscale problems, and is about to go from product engineering to program engineering as an industry.
Welcome to Program Engineering
medium.com/@jakejooyoung/wel…
Serroean retweeted
Apple just won the AI war
They announced the MacBook Pro M5 Max
What would have cost $40,000 to do with Nvidia chips, now costs $3,000 with a laptop
• 4x faster AI speeds
• Up to 128gb of unified memory
• Frontier intelligence on the go
The biggest complaint about Apple before was their AI speeds
You get insane memory to run big models, but the inference was super slow
Not anymore
Apple played the long game by not spending trillions on AI researchers. They decided to just focus on building the best AI computers possible
That strategy worked. It's over.
Readers added context they thought people might want to know
There is no general 4x speedup for AI. Apple only benchmarked the time to first token (TTFT), so only the prompt *processing*, but neither the reasoning/thinking, nor the response creation. And only a small 8B model in 4 bit quant.
See footnote 1 on apple.com/macbook-pro/ (archive web.archive.org/web/2026030314…)
Serroean retweeted
Our VC is adamantly opposed to this idea and I need your help.
I’m trying to push for “Co-programmable AI for Dev Teams” as our one-liner.
Our VC thinks it’s stupid.
Our idea is that there needs to be a way to co-program the AI dev teams use on a daily basis. Programming should happen live in-meeting through voice. Just like a coworker would learn the dependencies and what’s important to whom on the job.
Wdyt? Do I fold or double down?
Image is of our current one liner
There are levels to building software.
Most do 1 and 2 really well and have no idea what it takes to get to 6.
Mockup
Something that works on your laptop
Something that works for a single user reliably
Something that works for many users reliably
Something that works for many users with realtime interactions reliably
Doing all of this securely