We have the BEST anti-fraud models, folks - 99% precision, just like the B2s! Nobody beats that. Even Tiffany of @Reddit admits her team uses phone numbers to verify REAL users, but SADLY she’s stuck with @twilio - total disaster, terrible solution! She and her friends from other

Hey @v_tumas - saw Base Squelching is live in rippled v2.5.0. I’ve got a private XRPL on AWS with extensive metrics tracking and would like to test the basic version of the squelching algo in my setup. I’d be happy to submit performance & cost reports and loop you in on how it

We've got the 'k' in the number

Butler will make your dining decisions easier, faster, and more personal. (🧵) 🧵below Check it out on App Store 👉 https://t.co/VnGlvuzTsA

The new wonder

Truly realistic

If you're using @duckdb and @Minio, here's a nice tip on how to create a secret. Don't forget to modify region, the endpoint and the URL style.

"Architected for speed. Automated for easy. Monitoring and troubleshooting, transformed!"

Introducing Torso, a bimanual android actuated with artificial muscles.

a smooth start to one of the best visualization libraries out there

The friend is @taoolant

Tech wise I think he used @typesense for the search with embeddings using some embedding model - don't know which. I think @PostgreSQL for warehousing the reviews and restaurants and good old @nextjs to unite them.

Once you go to the second screen, you get a nice AI, Mr. Butler, who will talk to you and he has a massive set of reviews from various source (specifically pertaining to the restaurant of interest you had searched).

A friend has been building https://t.co/U5wCYcuZ63 and it is crazy good. Imagine it like a @perplexity_ai for food. You can search restaurants in human queries such as "pizza place in SF", "restaurants with outdoor seating and vegetarian options" or "lunch restaurants with good

How does one acquire such tikz powers?

Not all records you’re happy about breaking: @Cloudflare recently mitigated the largest ever reported hyper-volumetric #DDoS attack. 3.8 terabits per second (Tbps) and 2.14 billion packets per second (Bpps). Handled automatically any without any customer impact. Details to come.

Introducing Notebooks in Warp Drive Now your runbooks can live next to your command line. ◆ Compose with Markdown-flavored editor ◆ Embed Workflows or executable commands ◆ Sync updates to teammates in real-time

Prisma engineers when they come home after work

After this point, the conversion rate drops as the market reacts to you. Churn stabilizes, but 3% churn is still significant. Long term, I assume you can capture clients at a steady #ConversionRate but significantly less than the peak. And in the #LafferCurve spirit, at 100%

At some point, you have captured a portion of the market, putting you at the 🦄 rate. In the graph above, this is represented by point B. It is here where you are able to onboard users at a very rapid rate. Forgive my 2.5% on the graph, it is just a placeholder.