Jake Cusack (@Jakecusack) header
Jake Cusack (@Jakecusack) profile picture

Jake Cusack

@Jakecusack

MCSE | MCSD | MCSA | MCPD Azure & M365 Architect | AI Engineer | Data Scientist

 London, England
deepify.ai/
Joined July 2015
  • Tweets891
  • Following1.1K
  • Followers69
568Media
Photo by Jake Cusack (@Jakecusack)Photo by Jake Cusack (@Jakecusack)Photo by Jake Cusack (@Jakecusack)Photo by Jake Cusack (@Jakecusack)Photo by Jake Cusack (@Jakecusack)Photo by Jake Cusack (@Jakecusack)Photo by Jake Cusack (@Jakecusack)Photo by Jake Cusack (@Jakecusack)Photo by Jake Cusack (@Jakecusack)Photo by Jake Cusack (@Jakecusack)
StudyScribe is live on the App Store today. Record any lecture, lesson, or voice note. On-device AI turns it into a summary, flashcards, and a quiz. Built with Apple Intelligence. Nothing leaves your phone. If we call it private, it has to actually be private.
1
104
Three vendors, one day, zero secrets. Claude Code reads Entra oid, Copilot does app to app auth, Formal runs agents on prod with OIDC. Identity is the new secret. deepify.ai
18
Microsoft Research proves the 31-year-old KLS conjecture
1
11
— A math problem open since 1995 just got solved, and AI-guided search helped close it.
1
Databricks launches Agent Memory in beta
3
1
34
— A follow-up question that took eleven tool calls and nearly six minutes now takes two calls and forty-one seconds.
2
Microsoft Research proves the 31-year-old KLS conjecture
1
17
— A math problem open since 1995 just got solved, and AI-guided search helped close it.
2
Databricks cuts AI coding spend by up to 90%
1
10
— Databricks just told the story of its own AI coding bill getting out of control, and how it clawed costs back by as much as 90%.
5
Anthropic donates Agent Skills to the AAIF
1
10
— Anthropic just gave away control of one of its own standards.
2
OpenAI improves ChatGPT for low-resource coding tasks
1
9
— OpenAI is pushing ChatGPT's coding skills somewhere most benchmarks ignore: low-resource programming languages.
4
Databricks' MemCurrent triples agent pass rate with memory
1
22
— Triple the pass rate, using fewer tokens, running faster: that's what happened when Databricks gave its production agents a proper memory architecture instead of no memory at all.
4
GitHub Copilot adds a model decision router
1
14
— GitHub Copilot now makes a decision before it makes a single tool call: which model should handle this turn.
11
Microsoft previews Copilot-curated SharePoint knowledge base
1
38
— Copilot is learning to compile a team's email, messages and meetings into markdown files its agents read before they act.
8
Postman narrows 170 tools to 15 for its AI agent
1
5
— Give an AI agent more than about forty tools to choose from, and it starts picking the wrong one.
3
Load more