←
Paul Klein IV5,521 posts
Paul Klein IV avatar✓

Paul Klein IV ✓

@pk_iv

founder of @browserbase

◷ Joined October 2010▧ 528 Photos and videos
Paul Klein IV avatar
Paul Klein IV@pk_iv·Sep 14

this is the most important talk i've given in my time as a founder. in the next twelve months, we believe that every task that humans do online will be assisted by agent. watch this if you want to see the future that we're building for @browserbase

18:25
Samuel Spitz avatar
⟳ Paul Klein IV retweeted
Samuel Spitz@samuel_spitz·17h

Replit just launched DataCat. An open-source clone of the leading observability platform. We used a swarm of agents with @browserbase to test thousands of flows in the leading tool. Then rebuilt every flow 1:1. SaaS is dead, you can replicate anything with Replit Missions.

Post attachment
1:19
Paul Klein IV avatar
Paul Klein IV@pk_iv·15h

the number of meetings you have per week is inversely correlated to your effectiveness as a CEO. the weeks where i've been wall to wall in meetings are a symptom that i am the bottleneck. if i'm doing my job right, the team doesn't need me at all. tough pill to swallow!

Paul Klein IV avatar
Paul Klein IV@pk_iv·Nov 12

ThoughtSpot's CEO runs 160 agents across his company. He still won't let an LLM write a single SQL query. I sat down with @karkhanis to talk about why analytics has to be deterministic, why most AI pilots never make it to production, and what happens when agents read your data instead of people. 02:57 "Self-service was a hoax" 04:33 The insight hiding in Gong calls and Slack threads 09:28 Why the LLM should never write your SQL 13:41 "Data warehouses were built for humans, not agents" 17:44 Data lakes that became puddles 20:14 "No token anxiety" 27:32 Build vs. buy: "DIY is cute" Watch the full pod on YouTube, Spotify, and Apple Podcasts.

41:12
Paul Klein IV avatar
Paul Klein IV@pk_iv·Nov 12

Replit is a great example of a company that's "built the whole product". If your agent can write code, it can do anything (like shopping online). That's why we're super excited at @browserbase to be powering their browser agent capabilities!

0:15
Paul Klein IV avatar
Paul Klein IV@pk_iv·Nov 10

this might be the first black friday where agents buy more than humans

George Matelich avatar
⟳ Paul Klein IV retweeted
George Matelich@george_matelich·Nov 10

Over the last 5 months, usage of our platform has increased by 500%. And it happened because our customers loved the product, felt support from our team, and kept asking us to do more. When we launched Rely in January, we had three customers and one job: multifamily lease audits. We thought that was the company. Our customers broke that plan pretty quickly. Teams didn't stop at standard leases. They kept stretching the boundaries of what a lease audit was and could be. Then we shipped vendor contract audits and utility audits, and started getting asked for bespoke comparisons. As we met more operators across the country, we realized just how siloed their workflows really are: • Diligence teams trying to verify rent rolls and find buried risk before closing a transaction. • Transitions teams racing the clock to onboard thousands of resident records, contracts, and utility accounts without dropping a ball. • Operations and asset management teams trying to run ongoing compliance, financial, and lease audits across an active portfolio to catch errors before they turn into real losses. Every team was stuck re-doing manual work in spreadsheets because nobody had connected the dots. The takeaway was simple: nobody in multifamily wants five separate point solutions. They want one place to drop the data room, connect their sources, and know what's real. And they want to know the system was built specifically for multifamily teams. So that's what we built. Introducing the Rely platform. Whether you're closing an acquisition, taking over management, or managing an active portfolio, Rely connects your data sources and audits what you need: detailed lease terms, guarantor data, utility bill-backs, student housing pre-leases, vendor contracts, and compliance. And every single finding links directly back to the source document so your team can verify it in seconds. Over the last few months, we've processed millions of pages for teams like Cardinal Group, American Landmark, Olympus Property, Preiss, and SPM, taking audits that used to eat up days or weeks of manual review down to about an hour. Rely grew up fast because our customers wanted more. Check out the video below to see the new platform in action. If you're an operator, owner, or lender dealing with diligence, transitions, or portfolio audits at scale, we'd love to show you what we've built.

1:42
Paul Klein IV avatar
Paul Klein IV@pk_iv·Nov 9

fable 5.2 has some of the best computer-use capabilities we've seen

Post attachmentPost attachment
rahul avatar
⟳ Paul Klein IV retweeted
rahul@0interestrates·Nov 5

Today we're releasing DeepFinance, the first ultra-long horizon benchmark for high finance work. A set of 21 problems that reflect work at top PE and IB firms, DeepFinance is built using real deal data. Problems are constructed by experts with backgrounds from top firms like Evercore, JPM and UBS, and institutions like HBS and Wharton. Problems represent work that would take human professionals 20-40 hours to complete. Can AI really diligence a $100M+ deal? Finance benchmarks today are limited to evaluating agents on QA on financial docs or grading the numerical correctness of simple financial models. However, real world high finance work is full of assumptions and conflicting evidence. We built DeepFinance to realistically evaluate agents on those end-to-end finance workflows. DeepFinance leaderboard: juliusresearch.ai/deep-finan…

Post attachment
Paul Klein IV avatar
Paul Klein IV@pk_iv·Nov 4

you can now restrict what types of websites your agent can browse to using navigation policies.

Post attachment
Paul Klein IV avatar
Paul Klein IV@pk_iv·Nov 3

with the rise in all these cyberattacks, i have to imagine self-hosting/byoc is going to become table stakes for every managed infrastructure providers out there.

Paul Klein IV avatar
Paul Klein IV@pk_iv·Nov 2

i think people underestimate how big of a pain in the ass chromium is to run. it's one of the most attacked open source projects in the world, and changes that break playwright aren't uncommon. we patch and release chromium builds pretty much daily at this point.

ClickUp avatar
⟳ Paul Klein IV retweeted
ClickUp@clickup·Nov 2

Behind every office party hero is a browser agent doing the actual work. Right, @browserbase? 👀

0:33
Paul Klein IV avatar
Paul Klein IV@pk_iv·Oct 28

Your agents are blocked by data, not models. Last week I talked to @clickup's Head of AI @mathemagic1an, about why a dense context graph beats a smarter model every time. Their new Brain² Browser product uses @browserbase to fill forms, grab downloads, and bring external data into the platform. 07:45 What refactors taught clickup about narrow ai use cases 10:39 The biggest problem stopping companies from operationalizing ai 11:19 Why a dense context graph beats siloed data 12:09 Evals first: building the ai stack as a context engine 15:21 Skills as memory: clickup's answer to compaction 19:26 "I think we're close": are humans done writing code? 21:59 Ultracoding and the $120k bun to rust rewrite 24:13 What's actually stopping ai from curing cancer 26:30 Teaching an agent to order a pizza with just a browser Watch the full episode of Navigators on YouTube, Apple Podcasts, or Spotify.

38:02