Jesus Vargas

Founder and CEO · @phosailabs and @LowCodeAgency · Strategy first. Systems second. · 400 + projects · Zapier, Coca-Cola, Medtronic, American Express.

[Jesus Vargas @eltintero]

https://t.co/Tfl63vsWiJ

[Jesus Vargas @eltintero]

I used to spend an hour every morning just figuring out what was going on in my own company.

Scattered across Gmail, Slack, Drive, call recordings. The coordination tax nobody measures.

So we built 17 AI employees to kill it. Here's what happened 🧵

[Jesus Vargas @eltintero]

@CepnikMaciej it's not

[Jesus Vargas @eltintero]

8 traits that will define who wins in the AI business world:

1. Makes decisions when data is incomplete
2. Questions inherited processes
3. Uses AI to pressure-test ideas, not just save time
4. Moves fast from learning to doing
5. Translates technical capability into business outcomes
6. Leads with psychological safety through uncertainty
7. Seeks change before disruption forces it
8. Builds systems where humans and AI learn together

Most teams have 2-3 of these.

The ones with 6+ will be untouchable in 3 years.

[Jesus Vargas @eltintero]

The sequence I recommend before any build:

Map the operation. Two weeks understanding how decisions actually get made, where data lives, which processes run on one person's knowledge.

Document what exists, not what was supposed to exist.

Then build the AI Foundations, train the team and then install the systems.

400+ projects taught me if you skip any step, the next one does not work.

[Jesus Vargas @eltintero]

I keep getting the same question from founders this year. "How do I make what works on my laptop work for everyone I employ?"

Putting together something that answers it properly. Bringing in guests who have been inside that problem.

More soon.

[Jesus Vargas @eltintero]

The point isn't the tech. It's that we ran this on ourselves first.

When we tell a client "here's what AI employees do," we're not guessing. We've lived every workflow.

Case study: https://t.co/NxvGFmUvBw

[Jesus Vargas @eltintero]

I used to spend an hour every morning just figuring out what was going on in my own company.

Scattered across Gmail, Slack, Drive, call recordings. The coordination tax nobody measures.

So we built 17 AI employees to kill it. Here's what happened 🧵

[Jesus Vargas @eltintero]

Finance, marketing, HR — same treatment. 17 agents across 6 areas. Brevo, PostHog, Glide, Calendly, Typefully all wired in.

The numbers:
→ 20 hours/month back
→ 3x output
→ 1 hour of morning chaos → a 5-min read

[Jesus Vargas @eltintero]

Most companies use AI wrong.

They give it tasks. They't give it their problems.

'Write this email' is a task.

'Help me build a system that turns discovery calls into first drafts without touching my calendar' — that's a problem.

AI at the task level saves you hours. AI at the problem level changes your business.

[Jesus Vargas @eltintero]

Most companies are hiring for the wrong thing in the AI era.

They want: prompt engineers. AI tool experts. 'AI-native' workers.

What actually matters: people who can make decisions when the data is incomplete or contradictory.

That's the #1 trait that separates teams that win from teams that get disrupted.

Everything else is secondary.

[Jesus Vargas @eltintero]

I built @phosailabs because serious companies need more than a build. They need an AI strategy, a team that implements it correctly, and a partner that stays through the window where the P&L actually moves.

Most firms leave at go-live. That is exactly when the real work starts.

This is the practice I built to stay in that room.

[Jesus Vargas @eltintero]

Every founder I know has proven AI works on their own account.

Analyses that used to take a week. Proposals in their own voice in an hour. Contract clauses caught before the call.

My assessment when I hear this: they have already done the hard part.

The gap is making it work for everyone they employ. That is a strategy problem. The technology is downstream of it.

[Jesus Vargas @eltintero]

I built @phosailabs because serious companies need more than a build. They need an AI strategy, a team that implements it correctly, and a partner that stays through the window where the P&L actually moves.

Most firms leave at go-live. That is exactly when the real work starts.

This is the practice I built to stay in that room.

[Jesus Vargas @eltintero]

Every AI implementation we have built starts the same way.

Before the tools, the automations or anyone writes a single prompt: we document the context, define who owns what, and agree on what good looks like.

That foundation is what makes everything else work. The tools are easy. Building something that runs the same on day one as it does a year later, with one person or twenty, that is the part worth getting right.

[Jesus Vargas @eltintero]

@FabGeorjon

[Jesus Vargas @eltintero]

After 400+ implementations, the pattern that determines whether AI actually works in a business is always the same.

What determines success is the quality of context you give it before you ask it to do anything.

The tool is 10% of the equation. The documentation, the voice profile, the process playbooks that's the 90% most companies skip because it feels like setup rather than progress.

It is progress. It's the only progress that compounds.

[Jesus Vargas @eltintero]

@glideapps @workwithsepy no more testosterone for you mr Sepy