anton lin

daydreaming | prev. founded @scoutapp_ai, now StockX

[anton lin @nton__]

We sat down with Joe Lonsdale (@JTLonsdale) and explained why “Dental Billing” is one of the most compelling & complex applications of LLMs.

- AI Phone calls to insurance companies
- vLLMs to read 300+ page pdfs
- Web agents to navigate insurance portals
- Reasoning models to dispute denials

At DayDream we are transforming American small businesses with AI & allowing dentists to focus on what they love – treating patients.

If you're serious about working on difficult problems, we're hiring for Eng/Ops/Sales roles!

DM me or see https://t.co/oRYSoFgXYS

[anton lin @nton__]

another day, another clerk downtime incident

[anton lin @nton__]

If anyone tells you parsing PDFs is a solved problem they’re either lying or have only ever parsed one PDF

[Kushal Byatnal @kushalbyatnal]

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.

[Shreyas 🦷🦷🦷 @sparab22]

You could pay for fake AI talking head UGC, or you could just do a good enough job that they talk about you for free.

We could just be doing it wrong? 🤷 idk

[Harrison Kinsley @Sentdex]

Apparently an unpopular opinion, but I don't think Anthropic owes anyone heavily subsidized tokens for their third party app.

[anton lin @nton__]

@namanambavi congrats naman!!

[anton lin @nton__]

insane X axis…

[Cline @cline]

GPT-5.4 also has a 1M context window, but their evals show that needle-in-a-haystack (MRCR v2) scores 97% at 16-32K tokens, drops to 57% at 256-512K, and just 36% at 512K-1M.

So it's a good idea to compact regularly!

[anton lin @nton__]

The only time I use Wispr Flow is when I'm eating wingstop and making v0 mockups at the same time. This is true flow state.

[anton lin @nton__]

@calvinchen @MatternJustus @navidkpr Huge! Congrats Calvin

[anton lin @nton__]

@aadilpickle Anything interesting on the graph over time in the app?

[anton lin @nton__]

@sparab22 @DougMcMillon Insane ice on the wrist

[Sam Lambert @samlambert]

simple engineering is better engineering. everyone is over complicating everything.

[anton lin @nton__]

@0interestrates @juliusai whos the model 😍

[anton lin @nton__]

@nateberkopec @scoutapp_ai @whop Whop is in great hands… and @cultured is a killer

[anton lin @nton__]

this unironically solved half of our performance issues at @scoutapp_ai back in the day

[Nate Berkopec @nateberkopec]

If you have not put jemalloc on all your production Ruby systems yet, you should probably clear your schedule.

[anton lin @nton__]

@nateberkopec @scoutapp_ai aah the good old days.. hopefully you use us as a good example 😛

Currently trying to help dentists get paid for their work at https://t.co/CDF3hPLpCp

[anton lin @nton__]

@annietma you can book your appointment with me any time 😌

[anton lin @nton__]

HQ2 coming along nicely

[anton lin @nton__]

@CompositeAI @KicksCharlie let’s go!! @KicksCharlie @YangFanYun

[anton lin @nton__]

@rycao18 @66_shrey 's baby

[anton lin @nton__]

🦋