Extend

Over 1 billion PDFs are created every day, but your agents still can’t read them reliably. Today we’re releasing Parse 2.0, the most accurate document parsing API in the world. Extend already processes millions of pages daily for leading AI teams like Brex, Mercury, Opendoor, Flatiron Health, and hundreds of others. Now, its even better. Parse 2.0 is SOTA quality on RealDoc-Bench, our open source benchmark that measures agent success rate on real world docs that agents actually encounter in production. We trained Parse 2.0 on 1M+ pages of the hardest documents seen in production. Here’s how it stacks up: - #1 in healthcare, real estate, logistics, and financial services - 95.7% agent Q&A accuracy on 581 docs (next best: 92%) - 0.847 F1 on layout (next best: 0.759) Give it a try today and build production-ready document agents with Extend.

At @ExtendHQ, Jordan Meyer helped build the technical foundation as one of the company’s first engineers. When the team evaluated billing infrastructure, the decision came down to speed, flexibility, and long-term leverage. “Using Orb cut our billing project from 6 months to 3 weeks.” That’s the kind of engineering mindset we love supporting.

👀

earlier this year, we hand packed 65 gift boxes for our customers each box had cashmere gloves, scarves, and beanies, wrapped in tissue paper and tied with a blue ribbon by hand the responses from our customers made the effort all worth it 🫶

the kind of feedback that makes hosting so rewarding 🥰

a NYC rooftop bar on a warm april evening is undefeated as an event venue

something I've noticed — the gap between functional UI vs. UI that leaves an impression comes down to taste and craft...and you can't teach those things (nor can you copy them) it's what sets companies like Linear, Figma, and others apart that's what great design engineers bring to a team @andrewlu0 shipped this timeline replay view in Extend (to make debugging multi-step pipelines easier), and the little details just take it to the next level

Made in NYC with AWS: Kushal Byatnal, CEO & Co-founder, @ExtendHQ. Byatnal talks building document processing pipelines and building a successful startup in New York. From migrating to the city to migrating to AWS, the founder discusses the need for speed and the importance of partnership.

extend turned 3 last month! between the product releases, new hires, and closed deals, things move fast here. but the milestones are worth celebrating :)

try it out in playground! dashboard.extend.ai/studio/parsers…

new: conditionals and validations in edit schemas! you can now define rules like - "if this checkbox is checked, this field is required" - "only one of these options can be selected" directly in your edit schema. this is useful for complex forms where fields depend on each other, like healthcare intake, insurance applications, or tax documents

try it out in our playground: dashboard.extend.ai/studio/parsers…

new in extend studio: you can now compare parser runs side by side. change a config, rerun, and see exactly what changed in the output! useful for comparing things like - agentic table mode on/off - advanced chart extraction on/off - parsing engine version

comment if you want one of our retro stickers :)

vibe coded a new version of @ExtendHQ's website & learned that tragically, most mobile devices do not support comic sans by default live @ extend.ai/docs!

turned our website into a interactive google doc for April fools only felt appropriate (and threw in some hidden easter eggs everywhere)

splitter playground: dashboard.extend.ai/studio/split/p…

we just shipped a major speed upgrade to our document splitter 🚀 a customer ran a 1,044-page file through it with 315 (!!) different document types. it finished in under 5 minutes, at an average of 0.27 seconds per page. try the splitter for yourself! link below.

link to case study: extend.ai/resources/nuvo…

we're excited to share our work with @nuvocargo! nuvocargo partners with extend to process complex cross-border freight documents for a network of 30,000+ trucks. using extend's classification and extraction APIs, they hit 97-99% accuracy across document intake, classification, extraction, and shipment attribution, with near zero human involvement. nuvocargo has one of the most sophisticated AI/ML teams we've worked with, and it's been a pleasure partnering with them. link to the case study in the comments.