Andrey Fateev | Tech PM | Money, Systems, AI

Built 50+ Projects and Businesses. A few made real money.
I share what worked, what failed, and the systems behind it.
AI | B.S. Petroleum | M.S. Analytics

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

Big update: I’ve launched https://t.co/gWPkvjsgK5.

We build automation systems that connect tools, reduce manual work, and improve operations using Make, Shopify, APIs, AI, and analytics.

I’ll share the systems we build, the problems they solve, and what actually works.

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@MegaGeopolitics Tobleroneeeeeee nooooo

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

Crypto is ripping again.

BTC is up 20%+ this week.
ETH is up around 25%.

Now circle August 27–29.
Jackson Hole this year is literally focused on:

“Financial Innovation: Implications for Payments and Policy.”

With crypto already moving this hard, I wouldn’t be surprised if this becomes the next major volatility trigger.

Up or down?

I’m leaning toward a pullback.

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@ItsBitcoinBruh ?

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@sattyyouneed 444

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@elonmusk Optimus: “Grok, that was the wrong kidney.”

Grok: “You’re absolutely right. Sorry about that. I’ll redo it.”

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@MMAFullMount 😅😂

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

🚨 Vibe coding breaks at the first million users.

Not because the code is necessarily bad.

Because scale introduces an entirely different class of problems:

🚦 Rate limiting
🧠 Cache invalidation
⚖️ Load balancing
🗄️ N+1 queries and database indexes
🔌 Connection pooling, replicas, sharding
📬 Queues, pub/sub, retries, exponential backoff
🔁 Idempotency, circuit breakers, saga patterns
⚔️ Race conditions and deadlocks
🔒 Distributed locks and eventual consistency

Then one day the app goes down at 4 AM.

Now you need:
📊 logs
📈 metrics
🧵 distributed tracing
🚨 alerts
🎯 SLOs
💸 error budgets

Just to answer one basic question:

Why did p99 latency triple, and what exactly should we fix?

And deployment gets harder too:
🐳 Docker
☸️ Kubernetes
⚙️ CI/CD
🧱 schema migrations
🚩 feature flags
🔵🟢 blue-green releases
🐤 canary releases
↩️ rollbacks
❤️ health checks

Shipping a feature is easy.

Shipping it in a way you can safely roll back is engineering.

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@_MegaPolitics i dont think so

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@GBC_Press thats big

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@WallStreetMav thats just a number, relax and watch

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@danellisona Our pre-revenue is about 1trillion siiiir

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

Hardware!!!

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

The AI coding boom created one funny problem:

Now almost everyone can build.

Far fewer know how to sell.

Code is becoming cheaper.

Distribution is becoming more valuable.

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@elementfatron nahh😅

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

#ETH moved up hard today.
Are you bullish or bearish for the next 7 days?
I think we get a pullback within the next week📉

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@CoinMarketCap No

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

Yesterday I mentioned that we’re building our own internal intelligence system for the Texas water-well market.

Part of the reason comes from a business we tried to buy last year.

Let’s call it Company X.
It had been around for decades, had real equipment, real revenue, and a recognizable local name.

But due diligence changed how I look at buying small businesses.

1. 🏢 What exactly are you buying?
The history was spread across different entities, names, licenses, an Inc., a trust, and different operators.

A business can say it has decades of history.

But if that history doesn’t cleanly belong to the entity you’re acquiring, how much of that goodwill are you actually buying?

2. 👥 Where are the customers?
There was no proper CRM.
No clean customer history.

No easy way to answer:
Who was the customer?
What did we do?
When did we do it?
How much did they spend?
When should we call them again?

A phone full of texts is not a customer database.

3. 📈 Where does the revenue actually come from?
One year looked strong.
But when we broke it down, a huge part of that year came from service work.

The next year, service revenue dropped hard.

Meanwhile, drilling revenue followed a completely different pattern.

That matters.
You’re not buying last year’s revenue.
You’re buying whatever can realistically happen again.

4. 💰 What is normalized profit?
Reported profit is one number.
Normalized profit is the one I care about.

We found questions around COGS, payroll, depreciation, service volatility, and other operating expenses.

Before valuing the business, I built low, base, and high scenarios instead of trusting one reported year.

5. 🚜 Are you buying a business or expensive equipment?
Rigs, trucks, casing, pumps, and inventory all have value.
But equipment alone is not a business.

If the customers, processes, data, and relationships leave with the owner, you might not be buying a functioning company.

You might just be buying a very expensive collection of assets that now need fuel, maintenance, repairs, insurance, storage, and people to operate them.

At that point, some of those “assets” start looking a lot more like liabilities.

6. 🤝 What happens after the owner leaves?
This one is underrated.
For a business like this, I wouldn’t want the seller handing over the keys on Friday and disappearing on Monday.

We included a formal transition period in the deal structure.

Customer relationships.
Vendors.
Employees.
Technical knowledge.
Regulatory knowledge.

Some businesses need months of transition support before the buyer truly owns the operation.

That deal never closed.

But the due diligence was probably worth more than a lot of businesses I actually bought.

Don’t buy the story.
Buy the evidence.

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

A few of us from LSU had been talking about starting a water-well company in Texas.

With my petroleum engineering background, drilling itself was the obvious way in.

Then we found a more interesting angle.

Now we’re building our own internal intelligence system around the Texas water-well market.

850K+ records so far.

Depths, aquifers, drilling history, licenses, contractors, geography.

Let’s see where this goes.

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

Building software is getting easier.

That’s why I think the bigger opportunity is connecting software to a real physical business.

Not SaaS for the sake of SaaS.

Use software to understand the market better, choose where to go, who to work with, what to build, and how to operate smarter.

Things that used to require a team can now be built in weeks.

That’s the real leverage.

👉 Want a couple of examples?

[Andrey Fateev | Tech PM | Money, Systems, AI @andreyfateev77]

@WatcherGuru nice