Kushal Byatnal (@kushalbyatnal) on X (original) (raw)
Kushal Byatnal
Kushal Byatnal

update: the blinds are all closed this morning


our office is directly across the street from Mamdani's campaign HQ we looked over after standup today and surprise, Elizabeth Warren and Mamdani in a deep debate team is tempted to put up a sign petitioning for
@ExtendHQ
to take over doc processing for NYC
our office is directly across the street from Mamdani's campaign HQ we looked over after standup today and surprise, Elizabeth Warren and Mamdani in a deep debate team is tempted to put up a sign petitioning for
@ExtendHQ
to take over doc processing for NYC

We just hit a milestone in document processing — 91 page scanned PDF w/ nested tables, handwriting, and complex layout. 10,400 data points extracted with 100% accuracy.

Klarna using AI to rip out Salesforce and Workday is pretty magical at first glance.... but I've also seen this before: - company sees 7-fig Datadog bill - kicks off internal build to "save millions of dollars!" - staffs up team of eng - 6 months later, realizes their mistake
Replying to @kushalbyatnal
the true cost of software is not the build, it's maintenance + opportunity cost day 1 of a new project: 0 tech debt, team is motivated, engineering spec is clean. lots of momentum! day 100: backlog of requests, codebase littered w/ hacks, team moved onto other priorities
Replying to @ThingsWork and @mgsiegler
Here's another perspective...crazy how fast light travels, and saddening that it's still too slow for interstellar travel

PDFs suck. We just raised $17,000,000 in funding to fix this problem once and for all. Extend is building the modern document processing cloud. See how Brex, Square, Checkr, and Fortune 500s use it to process millions of documents:

Replying to @kushalbyatnal
and any good engineer will tell you, that's always the hard part of software code is cheap, outcomes are expensive
Replying to @kushalbyatnal
does AI change that? yes, the cost of writing code is → 0 but... these costs haven't changed: - determining which code to write - opportunity cost of prioritizing it - figuring out how to effectively deploy it to users - maintaining and improving it over time


Introducing Composer — the first AI Agent for document processing. Get to production-grade accuracy, autonomously in minutes. In our early beta, some teams hit 99% accuracy on complex document tasks in under 10 minutes. Composer is an agent built to optimize schemas the same

personal update — I'm excited to announce
@ExtendHQ
! We're building an AI-powered operating system for unstructured data. 4 months ago, we went through
@ycombinator
's W23 batch w/ just an idea. Today, we're growing crazy fast and bringing AI to modern enterprises everywhere 📈

big milestone for Extend! our new website captures months of learnings and focuses on one core mission — helping technical teams transform complex documents into reliable, high-quality data

we just signed a 6-figure contract with a publicly traded company...that's been around since before PDFs even existed
Don't miss what's happening
People on X are the first to know.

