I rebuild the growth frameworks PE operators trust with AI, then run them on real deals in public.
Every week in Fully Deployed:
- Framework rebuilds on real portcos
- Add-on deal teardowns
- Why boring businesses win the AI era
Free: https://t.co/pK9F40AlVR
This post is indicative of a major problem with AI news:
There's so much earth-shattering tech being released daily that nobody can keep up with it anymore.
Brands now have to remind you. "Oh btw, we solved audio generation a few weeks ago. Just making sure you saw!"
40% of office workers say their job is pointless.
Finding those jobs takes 6 weeks and a team of consultants.
Plek AI (YC F26) finds them and builds the restructuring plan in minutes.
Founder mode for every company.
https://t.co/pvf9lJPdq6
RWS closed its Acolad deal and plans to push its AI language platforms into Acolad's enterprise clients.
Here's the scorecard I'd use to tell whether that's working. π§΅
Only #3 and #5 move enterprise value.
#1, #2 and #4 tell you whether they will.
Most AI cross-sell decks report #2 and stop. Attach rate without margin is a vanity metric.
Apex Service Partners: HVAC, plumbing, electrical. 150+ locations, ~75 local brands, reportedly $3B+ revenue. Apollo just bought in.
At that scale the edge isn't buying more shops.
It's call-center conversion, tech capacity, and onboarding the next one faster.
The quietest money in PE: the installed base.
Genomics instruments. Lab infrastructure. Surface-prep equipment. Diagnostics.
Sell the machine once. Sell service, consumables and upgrades for a decade.
Several deals last week were really about this. Few headlines said so.
California's new healthcare transaction rules took effect Nov 2.
Per McGuireWoods: PE funds and MSOs can trigger their own notice obligations, and some deals need notice 90+ days before closing.
If your deal timeline doesn't show this, your deal timeline is wrong.
No synergy estimates. No claims about private data.
Just a sharp question for a team about to build a stand-alone business from scratch.
That's how permissionless deal flow starts.
One Rock agreed to buy 7 coffee brands from Kraft Heinz for $650M EV. Closing expected mid-2027.
Here's the 2-page preview I'd hand the deal team, built from public info only. π§΅
Page 2: Three KPI trees.
1. Gross margin after trade spend
2. Distribution velocity by SKU
3. Working-capital conversion
Each with the data needed to prove it and the person who owns it.
Salesforce agreed to buy Cresta, a voice-agent company.
Every portco will get pitched AI voice agents this year.
The only scorecard: cost per resolved case, first-contact resolution, retention. After software and implementation costs.
"Agents deployed" is not a KPI.
Ethos reportedly made 3.8x gross selling Motocaddy, an electric golf caddy maker, to Corbel.
The next leg is North America.
The cheapest way to scale a niche brand: other people's channels. Dealers, service networks, accessories.
Own the brand. Rent the reach.
Kraft Heinz is selling its North American coffee brands to One Rock for $650M EV.
The detail I'd watch: Kraft keeps 30%.
A seller with skin in the game changes the separation. It's harder to leave stranded costs in a business you still partly own.