Ryan Bednar retweeted
What's super clear to me downstream of widespread superintelligence:
1. Startups will require less people to get off the ground.
2. There will be more startups.
3. Startups will become fundamentally more ambitious about the problems they solve.
An incredibly optimistic time. Don't overthink it.
Ryan Bednar retweeted
The execution velocity of @mynameisyahia and @getcontextdev is just insane. I donโt understand how itโs possible.
5 of our portfolio cos are now solving a key piece of our internal OC tech stack
- @mastra for agent framework
- @triggerdotdev for background agents and cron
- @getcontextdev for realtime web data
- @sazabi for observability and fixing slop
- @superset_sh for ADE
and more on the wayโฆ
what a time to be an investor right now.
Our thesis on backing @getcontextdev:
1. Smarter models will still need fresh context.
2. Yahia is a baller.
Context.dev turns the live web into structured data via one API, so developers can build agents (and not scraping infrastructure).
We raised $9M to give your agents the internet.
Led by @generalcatalyst, with @orangecollectv and Robinhood Ventures.
Search, scrape, crawl, monitor with one API trusted by 800+ customers.
Weโre making it ridiculously easy to build with web data.
Ryan Bednar retweeted
Introducing overshoot, it will cut your GPU bill in half with zero quality loss, or you donโt pay.
Over the past few months, we have forward deployed into AI teams to optimize their inference stacks. They have now cumulatively saved over $5M.
Overshoot is software that bakes in our hard-earned inference learnings into a scalable solution.
overshoot.ai
Ryan Bednar retweeted
Today, we're launching Sazabi version 1.0.
The reality of shipping software as a team in the age of AI is actually crazy.
Agents merging their own PRs. Hundreds of deployments per day. A codebase that is constantly changing. Junior developers with 10x or 100x leverage. ๐ตโ๐ซ
I spent the last year building a safety harness for teams that ship with AI. It's called @sazabi.
The premise is simple:
1. You send us logs.
2. We detect anomalies, investigate them, and send you back rich, actionable alerts about real problems in your app.
Sazabi helps you find issues before your users notice.
No monitors or noisy alerts. Just pure signal.
And today we're making it available for anyone to use.
Ryan Bednar retweeted
Coding this weekend? Need a web data API?
Your first 10,000 requests are on us
For the next 6 hours, i'm giving away 10,000 free credits on context.dev
Thatโs enough for any one of these:
> 10,000 scrapes, browser rendering + proxy escalation included
> 10,000 web searches
> 1,000 answers
> 1,000 brand lookups
> so much more
Comment what youโre building + DM me, Iโll add the credits myself!
AI has gotten unbelievably good at coding over the past year. @HalluminateAI is helping make it incredibly good at knowledge work next. Super bullish on this team!
Today we're excited to announce @HalluminateAI's $30M Series A led by @oakhcft with participation from @YCombinator, @orangecollectv, @FTPartners, @heavybit, and more.
Ryan Bednar retweeted
***3 days until our final close of OC Fund I!***
LPs often ask us how we actually select Y Combinator companies.
We have a super power: OC Portal
We manage everything in our pipeline, GP notes, due diligence, selection, portfolio tracking, and LP communication OC Portal. No Portal, No OC. Screenshots below.
For many of our LPs, access to Portal becomes their super power as well - because they can share it with their entire team. There is simply no better tool out there to invest into Y Combinator, than access to OC's software.
With Portal organizing us, we get to ask all the right questions:
๐๐ ๐๐ฌ๐ค ๐ฐ๐ก๐๐ญ ๐ข๐ ๐ข๐ญ ๐ ๐จ๐๐ฌ ๐ซ๐ข๐ ๐ก๐ญ.
We try to emulate PGโs amazing ability to see the most amazing future for a company. If weโre wrong, itโs fine - loss of 1x. But if weโre right - 500x can occur.
๐๐ ๐๐จ๐งโ๐ญ ๐ญ๐ซ๐ฒ ๐ญ๐จ ๐ฉ๐ซ๐๐๐ข๐๐ญ ๐๐ฎ๐ญ๐ฎ๐ซ๐ ๐ญ๐๐๐ก๐ง๐จ๐ฅ๐จ๐ ๐ฒ.
The best founders create markets. We want to learn from them what's they've figured out so far.
๐๐ ๐๐๐ญ ๐จ๐ง ๐ญ๐ก๐ ๐๐จ๐ฎ๐ง๐๐๐ซ๐ฌ ๐ฐ๐ก๐จ ๐๐ซ๐ ๐๐๐ซ๐ง๐๐ฌ๐ญ ๐๐ฎ๐ข๐ฅ๐๐๐ซ๐ฌ.
They will discover a market where none exists. And we're okay betting on a founder knowing his or her best ideas are still to come.
๐๐ ๐ฌ๐ญ๐๐ซ๐ญ ๐ฐ๐ข๐ญ๐ก ๐ก๐จ๐ฐ ๐๐๐ฌ๐ญ ๐ข๐ฌ ๐ ๐๐จ๐ฆ๐ฉ๐๐ง๐ฒ ๐ฌ๐ก๐ข๐ฉ๐ฉ๐ข๐ง๐ .
We're using of the actual product and we watch just how fast they iterate. It really is the best predictor.
๐๐ ๐ญ๐ซ๐๐๐ค ๐ฌ๐ฅ๐จ๐ฉ๐ ๐จ๐ฏ๐๐ซ ๐-๐ข๐ง๐ญ๐๐ซ๐๐๐ฉ๐ญ
So fast growth in tiny numbers blows away medium fast and large numbers. We don't ever look at absolute numbers like revenue at point in time.
๐๐ ๐๐๐ซ๐ ๐๐๐จ๐ฎ๐ญ ๐ฌ๐ข๐ณ๐ ๐จ๐ ๐๐ฆ๐๐ข๐ญ๐ข๐จ๐ง.
So formidable founders who are incredibly ambitious (Phil from Starcloud and Nico/Emily from Corgi) can beat out companies with zero revenue.
๐๐ง๐, ๐ฆ๐จ๐ฌ๐ญ ๐ข๐ฆ๐ฉ๐จ๐ซ๐ญ๐๐ง๐ญ๐ฅ๐ฒ, ๐ฐ๐ ๐ฐ๐จ๐ซ๐ค ๐ฐ๐ข๐ญ๐ก ๐จ๐ฎ๐ซ ๐๐ ๐๐ฅ๐ฎ๐ฆ๐ง๐ข/๐๐๐ฌ ๐ญ๐จ ๐ฌ๐๐ ๐ฐ๐ก๐ข๐๐ก ๐ฉ๐ซ๐จ๐๐ฎ๐๐ญ๐ฌ ๐ญ๐ก๐๐ฒ ๐ฅ๐จ๐ฏ๐ ๐ฎ๐ฌ๐ข๐ง๐ .
Most of our portfolio use each otherโs products. And most of our portfolio founders ended up investing in the fund as an LP. So the loop continues.
~~~
All of that leads to 15 companies who we've already done the due diligence on. Already used their product (when possible). Already met their customers. Already become their customer (when possible).
๐๐ง๐ ๐ญ๐ก๐๐ง ๐ฐ๐ ๐ฌ๐ก๐๐ซ๐ ๐ญ๐ก๐๐ญ ๐ฅ๐ข๐ฌ๐ญ ๐ฐ๐ข๐ญ๐ก ๐จ๐ฎ๐ซ ๐๐๐ฌ.
We tell them, ALL these companies are amazing. You should invest in as many of them as possible directly. We're about to pick four, we'll let you know which once we do. But the full 15 are amazing.
๐๐'๐ซ๐ ๐๐ฅ๐ฆ๐จ๐ฌ๐ญ ๐๐ฅ๐ฐ๐๐ฒ๐ฌ ๐ญ๐ก๐ ๐๐ข๐ซ๐ฌ๐ญ 7 ๐๐ข๐ ๐ฎ๐ซ๐ ๐๐ก๐๐๐ค ๐ญ๐ก๐๐ฒ'๐ฏ๐ ๐ ๐จ๐ญ๐ญ๐๐ง.
We say yes almost immediately and they yes almost immediately.
If you want more detail than that...gotta join as an LP. Accredited investors only.
And only 3 days left!
We made a bet at @Halluminate in mid-2025 that Finance would be the next big field after coding that would really take off from agent usage. Few reasons for this:
1. Finance is the world's largest knowledge work industry
2. The outputs are semi-verifiable (which has some nice properties for RL)
3. Financial reasoning generalizes to a lot of other knowledge work domains the same way SWE agent capabilities generalized to a lot of other engineering domains (ex. CAD, MLE).
Almost all my peers in banking have become increasingly AI-pilled in how they do work (similar to SWEs in Dec 2025 with Claude Code) and there's so much more progress to come soon.
Its even more exciting to see our hard work in RL Envs/Benchmarking play out in the hands of consumers.
If you'd like to contribute to building Financial Superintelligence we're hiring!
Today is the day that AI passed the "financial modeling Turing test" for me.
A "push button" build from scratch Skill that one-shotted a model on $MU that is indistinguishable (to me) from a model that a junior analyst would build from scratch. With full, impeccable adherence to all aspects of who i like to format, design & build models (took a few turns to dial it in).
Opus 5.5 is unbelievable.
Ryan Bednar retweeted
Introducing Superset for iPhone.
Remotely control agents running on your computer. Review and ship PRs from your pocket.
Installed Omarchy on a $325 Beelink SER3 mini-pc. Takes a day or two to learn the shortcuts and window bindings but after that itโs stupid fast. Lives up to the hype and then some.
Only one minor complaint: my iPhone still prefers macOS.
Somebody please make an Omarchy phone! @dhh
























