I wish there was a better term than "software factory" -- it's kind of uninspiring and dehumanizing and not really accurate in terms of the importance of humans in creating software going forward. But "factory" does seem to be the term everyone is settling on for automation
44
3
89
12,376
Kency retweeted
If you're interested in WebMCP, you should be following Idan Levin. He and his team have been doing extremely exciting work in this space- benchmarking, cataloguing, you name it! Strong suggestion to follow him.
Replying to @rauchg
You might find this interesting - We created the first WebMCP benchmark, comparing it with computer use and DOM-based methods across speed, cost, task success rate, and token usage. The results are incredible. webmcp.com/benchmark You’re getting 3-5x speed, and an order of magnitude cost improvement (up to 23x lower costs) while being 10-15% better at solving tasks compared to other methods
20
33
843
194,553
Kency retweeted
My latest deep dive looks at how @stripe, @Adyen, @Visa, @Revolut, @Plaid and @nubank are approaching transaction foundation models, where the technology works, where it breaks down, and the decision framework leaders should use before funding an in-house build. Most banks and fintechs should not rush to build one. Having millions of transactions is not enough. You need linked customer, account and merchant histories, clear outcome data, and enough value at stake when a decision improves. Stripe lifted card-testing attack detection from 59% to 97% with its Payments Foundation Model. Adyen says its model-led routing has increased conversion by more than 1%. Those are strong results. They also come from companies with deep data, fast feedback loops and infrastructure built to serve decisions in milliseconds. For many institutions, a hybrid approach will make more sense: • Build around behaviour that is unique to your customers and cannot leave your business. • Use external providers where network-scale signals matter, especially fraud and risk. • Combine both in a real-time decision layer that can approve, decline, route, price or send a payment for review. The key takeaway: transaction volume alone is not a data moat. Outcome data is. open.substack.com/pub/sambob…
3
4
61
5,507
Want to find AI PDLC for enterprise - any guide or instruction??
7
Cách tạo USB BOOT bằng DLC BOOT 2019 (không thể dễ hơn) nitter.net/blogchiase
Win a Shadow Daggers | Slaughter (Factory New) now! wn.nr/2StS4Z
#Giveaway #CSGO #CSGOGiveaway ★ Huntsman Knife | Damascus Steel (MW) Giveaway 😎 Win a free knife! Enter here: idle-empire.com/giveaway.php
18
1,465
98
Just entered the giveaway to win a NiP Jersey signed by @NiPGaming #CSGO Team! nip.gl/giveaway #GONINJAS @betwayesports
This is firsg time i try this stuff I hope it better more than facebook
2