A Gulf Coast refinery lost a process engineer in March, 22 years in the same plant. Nobody asked him why Unit 4 runs hot every summer before he left. That knowledge wasn't written down anywhere, and now it's simply gone.
English
Sensay ⛩️✨
2.5K posts
Sensay ⛩️✨
@asksensay
Streamline offboarding, capture knowledge, and preserve expertise.
Katılım Haziran 2023
302 Takip Edilen51.2K Takipçiler
The knowledge that keeps a factory running is mostly undocumented. Which machine sounds wrong before it fails, which supplier actually delivers, why a process exists at all. None of it is in the manual, and all of it leaves when the person does.
English
Sensay ⛩️✨ retweetledi
Sensay ⛩️✨ retweetledi
El CEO de Anthropic viendo como China lanza un nuevo modelo de IA que supera a Claude Opus 4.8 en TODO, iguala a Fable 5 costando 8 VECES MENOS, que encima es 100% Open source y no puede hacer nada al respecto
Kimi.ai@Kimi_Moonshot
Introducing Kimi K3: Open Frontier Intelligence 🔹 2.8 Trillion Parameters, 1 Million Context, Native Multimodal 🔹 Kimi Delta Attention enables up to 6.3x faster decoding in million-token contexts 🔹 Attention Residuals deliver ~25% higher training efficiency at <2% additional cost 🔹 Built for long-horizon agentic coding and self-evolving workflows Kimi K3 is now live on on Kimi.com, Kimi Work, Kimi Code, and the Kimi API. Open Weights by July 27, 2026. 🔗 API: platform.kimi.ai 🔗 Tech blog: kimi.com/blog/kimi-k3
Español
Visa and Mastercard are both building payment rails for AI agents. That tells you more about where this is heading than any token chart. When the incumbents copy the primitive, the primitive is real. Whether the tokens are is a separate question.
English
Tomorrow! Sensay Community Call with Dan Thomson
Join us on Friday 17th July at 2pm UTC for a Community Call and Project Update with CEO and Founder, Dan Thomson.
Watch Live on Youtube, Twitter or Linkedin.
Submit questions live in Chat or in advance via Telegram
youtube.com/watch?v=oLRJJ0…
YouTube
English
The Knowledge Ledger: How Web3 Can Rewire Enterprise Knowledge Management medium.com/p/the-knowledg…
English
Apple Lawsuit Exposes Enterprise Data Risks During Employee Offboarding - pymnts.com/apple/2026/app…
English
Real conversation from yesterday:
Prospect: "Can't we just have the departing employee train their replacement?"
Us: "How often do those timelines overlap?"
Prospect: "...almost never."
Exactly.
The traditional knowledge transfer model assumes:
→ You hire the replacement before the departure
→ They overlap for weeks of training
→ The departing employee is motivated to train thoroughly
→ The new person asks all the right questions
→ Nothing gets forgotten
In reality:
→ Hiring takes 2-3 months
→ You get maybe 1 week of overlap if you're lucky
→ Departing employees are mentally checked out
→ New hires don't know what they don't know yet
→ Everything critical gets forgotten
Sensay flips this. Capture knowledge once. Make it available forever.
The replacement can start asking questions on day one. Six months in, they're still learning from someone who left before they started.
That's how you actually transfer knowledge in 2024.
English
The knowledge people think is important vs what actually matters are two different things.
Managers want: Strategic insights, big-picture thinking, documented processes.
What new hires actually need: Where files are saved, who to ask for what, which clients are difficult, the 15-second workaround that saves 2 hours.
The mundane stuff. The tribal knowledge. The things nobody writes down because they seem too obvious.
Until they're not.
Sophia captures both. The strategic and the tactical. The documented and the invisible.
Because she asks follow-up questions. Probes deeper. Captures the context that Word documents miss.
One replica has been answering questions for 8 months after the employee left. 847 questions so far.
Mostly tactical stuff. The kind that would have been lost forever.
That's 847 moments where work didn't stop because someone left.
English
Three months ago, a Director of Operations gave notice at a mid-size manufacturing company.
23 years with the company. Knew every system, every workaround, every critical vendor relationship.
They gave him a Word doc to fill out. He got through page 2 of 47.
Fast forward to last week.
Their production line went down. The new Director couldn't figure it out. The vendor said it would take days to diagnose.
That knowledge walked out the door in June.
This isn't a rare story. It's every story.
$31.5 billion in lost productivity every year because we're still using 1990s knowledge transfer methods.
Meanwhile, the technology exists to capture, preserve, and transfer institutional knowledge at scale.
95% retention rate. $500/year. Available today.
The question isn't "can we afford it?"
It's "can we afford not to?"
English
McKinsey survey reveals nearly 1 in 3 European workers consider quitting. Poor offboarding costs firms up to $500K yearly in litigation, data loss, and brand damage.
ad-hoc-news.de/boerse/news/ue…
English
Hacked, Package
A Hacked npm Package Shows Why German Companies Can’t Afford to Ignore Employee Offboarding
ad-hoc-news.de/boerse/news/ue…
English
Problem: People hate filling out forms.
Solution attempt #1: Make better forms.
Result: Still terrible.
Solution attempt #2: Video interviews with real people.
Result: Too expensive, not scalable, people still guarded.
Solution attempt #3: Text-based AI chatbot.
Result: Felt robotic. Nobody wanted to type paragraphs.
Solution attempt #4: Voice-to-voice AI that actually sounds human.
Result: People talked for hours. Voluntarily.
Turns out humans have been transferring knowledge through conversation for 300,000 years.
Who knew.
Sophia now conducts 1,000+ exit interviews per month. Average conversation time: 47 minutes. Knowledge capture rate: 95%.
Compare that to your last Word document handoff.
Sometimes the best technology just lets humans be human.
English