Divyan

Building Auctra - content that's verified before it carries your name. 2 decades in strategy & ventures. Be known for what you know.

[Divyan @Divyan]

A month ago my wife's LinkedIn was dormant. Under 1,000 followers.

Then she started posting with Auctra.

28 days later: 21,745 impressions, 96% from outside her network - and she crossed 1,000 followers.

Same person. Different system.

Be known for what you know.

[Divyan @Divyan]

@elonmusk You got a robot driving a car that flies. This unlocks so many things at once. Congrats!

[Divyan @Divyan]

Access the underlying 89-market analysis here:

https://t.co/uNZoQslA0z

[Divyan @Divyan]

The national average almost hid the story here.

When we started going through the H-1B employment data, the US number was a 7.5% decline. Significant, but not extraordinary.

Then we looked city by city.

DFW was down 33.4%. San Francisco was up 86.7%. Raleigh was down 52%.

[GRAI @GRAI_AI]

75 of 89 US H-1B employment markets declined.

Nationally, employer-declared new-employment positions were down 7.5% between Apr-Jun 2025 and Apr-Jun 2026.

But that average hides an unusually wide spread across individual US markets.

[Divyan @Divyan]

Employment stress can eventually reach housing, but people sit between those two datasets. They have to decide whether to stay, move, sell, rent, buy somewhere else or wait. And another household may take their place. That human behaviour is where the real estate question starts.

[Divyan @Divyan]

The surprise here isn't the 33.4% DFW decline.

Sell a $650k US home once you're a foreign person for US tax purposes and FIRPTA withholds 15% of the sale price - $97,500.

Not your profit. The price.

Relocations get expensive when decisions run in the wrong order.

[GRAI @GRAI_AI]

You own a $650,000 US home.

If you're a foreign person for US tax purposes when you sell, the standard FIRPTA withholding can be:

$97,500.

Not 15% of your gain.

15% of the amount realized.

[Divyan @Divyan]

The costs above are the easy part.

The harder question is which checks to run first - legal access, permitted use, soil, drainage, approval timelines - and which ones quietly kill a deal.

We wrote up the full checklist: https://t.co/bnhpCiAdT8

[Divyan @Divyan]

A $100,000 plot of land can quietly become a $170,000 plot before a single brick is laid.

None of that shows up in the listing price.

[Divyan @Divyan]

Building @GRAI_AI, this is the part I keep coming back to: the most useful thing AI can do here isn't hand you a confident answer.

It's surface what's missing before you commit capital.

[Divyan @Divyan]

The next layer may be operational memory: the accumulated record of how the organisation actually behaved.

[Divyan @Divyan]

Google has agreed to pay $10M for part of Spirit Airlines’ internal data estate.

~100M emails across 80,000 accounts. 500M Teams messages. ~30M lines of code. Documents, spreadsheets, calendars.

Mercor bid $7.5M for the same archive.

A market is forming.

[Divyan @Divyan]

The more useful the structure, the harder anonymisation becomes.

We’ve spent years talking about proprietary enterprise data as an AI advantage, usually meaning documents and databases.

[Divyan @Divyan]

She used Auctra to do this. Start free, no card: https://t.co/xFPpB5NTRU

[Divyan @Divyan]

A month ago my wife's LinkedIn was dormant. Under 1,000 followers.

Then she started posting with Auctra.

28 days later: 21,745 impressions, 96% from outside her network - and she crossed 1,000 followers.

Same person. Different system.

Be known for what you know.