Chris Riley @LCSlates

my mornings... test, test, retest, test, retest, test, improve, retest, improve.

👀👀

Brendan (can/do) @BrendanFoody

7 hours ago

A research leader at a frontier lab told me the data market will grow by 100X. If Mercor maintains, let alone grows market share, we will be a trillion dollar company.

18 5 342 24K 179

this is awesome

Yahia Bakour @mynameisyahia

a day ago

“can you should stop launching on Friday nights, nobody’s even awake.” - @michael_chomsky Anyways, Introducing Batches scrape 100,000 URLs in one shot. give us up to 100K URLs, or just one and we’ll crawl from there. 9 output formats per page: > HTML > markdown >

26 7 166 10K 116

rl.data.cool

slightly ahead of schedule, business owners direct listings.

Chris Riley @LCSlates

2 days ago

the data complexities & transaction processing are so very challenging, but fun little milestones being hit hour by hour (after days of mind numbing pain - brutal). --> tech specs are live. --> workflow scores built and continue to be tested (synthetic testing right now). -->

0 1 6 2K 10

@danhoyes03 thanks Dan. its extremely complicated to get right (logic and execution), but would be amazing if I can get it just right.

Morning thoughts on a data marketplace... (rambling thoughts)... agreed, not a new concept, even in data. getting both sides together and actually transacting is the hard part, and relationships matter here. tech, ease of use, relationships.. all matter. the exciting aspect is there has never been a time like right now (my optimistic view point is on today). tailwinds at the 10k ft overview look strong. now, with my economics lens on, I look at how the market works today (and I see inefficiencies - like many growing markets): -Prices for the same kind of data run $10K to $5M, mostly unpublished -Buyers pay up to $50K just for an intro -90-120 days from first call to cash (longer sometimes) -many different layers/people/companies now in the mix of that transaction -The seller gets one quote and is paid once, even when the data can used again and again (changing, but in most cases) -The buy side is thin but growing fast: dozens of funded buyers, most of them new, with likely more coming (in my view very likely, but counter-argument is AI trade blows up in full). Just read something that there were 150+ buyers, half of which have been created in the last 2 years and the $$ they can spend (collectively) is unlike anything that has come before. For this specific data (there is plenty more out there btw - not just corpus data - but its the flavor of year), the sell side is thousands of companies that don't know what they have or whether they're one of the rare ones worth a premium. and the work to "supply" it; is substantial but small tech-enabled gains are happening to make this easier. still hard, still actually have to properly communicate, but gains are happening. TLDR; yes, a very hard goal to achieve (makes it exciting) there has never been a time like this (tech, $, growth; tailwinds) I'm happy to connect with all sides. could I be wrong? could this be dumb looking back 1 year from now? sure.... but i also could be right.

@TrentKrupp this is a fantastic idea. have you seen a few of these live already?

@Shpigford gl on surgery!

There is just no way things don’t work out. I’ve never seen a future so bright. All doom is fake. Things are amazing and only getting better. I’ve never seen more people winning, in every conceivable way possible. Losing shouldn’t even enter your conscious

@michael_chomsky 🔥🔥🔥

👀 Seeing plenty of this as well. This is actually hilarious (sad); he's now updated the image to to reflect 2026 (it was 2024), same month and day, but still in test mode, and "arrived on" also shows the year now of 2026. "o shit, dear opus, make an update this fake screenshot to reflect a 2026 transaction, make no mistakes, hurry"

Michael Fanous @michaelfanous1

2 days ago

For anyone following the data market here, I wouldn’t take every claim about buyer relationships or completed deals at face value. There’s a lot of misrepresentation, and we should lead with more transparency. The pool of legitimate buyers is very small. We’ve disqualified over

3 2 62 4K 71

the data complexities & transaction processing are so very challenging, but fun little milestones being hit hour by hour (after days of mind numbing pain - brutal). --> tech specs are live. --> workflow scores built and continue to be tested (synthetic testing right now). --> 3rd party verification talks continue. --> data transit testing continues to work in a series of more and more complicated corpus tests (health, PII, multi-modal) --> up to 115 listings by end of weekend, on the front end, we will have a direct company listing option.

Chris Riley @LCSlates

2 days ago

Morning thoughts on a data marketplace... (rambling thoughts)... agreed, not a new concept, even in data. getting both sides together and actually transacting is the hard part, and relationships matter here. tech, ease of use, relationships.. all matter. the exciting aspect is

3 3 24 2K 34

@jessiicawu agreed, but important note, unburdened by priors can be a negative if/when markets adjust, change, evolve (they always do). perfect combo; young and hungry + mixed with some seasoned vets with scars/war stories to pass along.

pray for me. need to hold down the vomit. the pup ate out his stitches (again) and while examining how many he ate out (this is his back of leg, thigh area), I get a deep whiff of anal sack. literally the worst smell on the planet. coming up... new stitches + *glands expressed. *by far the worst part of dog ownership.

🫡 - agreed all around and thank you. it could work, just not sure it will work (is it the right time, do I have the right logic, do I have the right relationships, do people even care or want to talk about this now). - if I can talk to both sides (daily), I think I can build something (with their help) that does work.

rings so true, everything you just said on both sides: --> "buy side looks liquid until you try to fill a tight spec, then every archive is the wrong shape or the rights aren't clean". spot on, and its a real hurdle to get over for the buyer. because its complicated (transactional and processing inefficiencies). it has to be vetted, trust-worthy, and hopefully a more much seamless process than what has transpired to date. Throwing X out because this is a bubble, but its generally true still, more entrants (sometimes bad entrants) to the market has made transacting a little more complicated actually. on the marketplace tech side, I spend 90% of my time on the data processing & transactional inefficiencies. how to constantly improve there on both sides (buyer and seller). very hard task - but I want to improve and build from what I hear directly from both sides. -- sell side has the opposite problem; waiting on a buyer who can actually decide. Also true and also a dynamic of new entrants flooding the market (good and bad). Imagine how many referrals micro1, mercor (plenty more we can list off) receive and shift through daily now. I can't even imagine. I'd venture a guess that 95% of these will never amount to anything (not even first steps to a deal). So there issue is supply is there, but is it the right supply? right deal? right counterparty? I'm sure they will figure it out -- $$ helps -- but that might not be the case for all buyers. **And so many layers in between (upstream and downstream) that can be talked about as well! all moving at the speed of something never seen before... so its the ultimately fluid market; hour to hour changes.

1. very true. i think this segment of the buyer side will be challenge to get in. but, i still think (if executed correctly), maybe longer-term it could even eliminate the pain points they see currently. however, with their capital stack, could they just raise and allocate more funds to a team there; sure. definitely a challenge for the bigger companies with huge $. 2. exclusivity is the big one I've seen. "shift the market, hence you control the supply and therefore the labs have to change how they think about data acquisition." --> that's the dream I'm building towards. Will it work? I hope so, but a challenge for sure!

@sk_2496 it varies day to day right now (hell, even hour to hour). in the marketplace directly, i'd lean to the buy side as the challenge. But, I can also confirm from the buy side, there are real issues/pain points (not surprising I guess) for where the market currently is.