Advisory & Investment

Mahmoud Saadat · Founder

I've spent my career
building hard things.

From semiconductors and automotive systems to autonomy, imaging radar, and software-defined vehicles. Today I advise and invest in early-stage Physical AI founders navigating the same hard problems.

No pitch deck required.

Physical AI is hard.

Not because the AI is hard.

Because physics, hardware, software, capital, manufacturing, and customers all have to work at once.

The Job

Build. Ship. Scale.

1

Build the right thing.

Architecture. Sensing. Compute. Power. Hardware/software boundaries.

Don't optimize the wrong system.
2

Ship it.

Validation. Manufacturing. Reliability. Cost. Supply chain.

A demo isn't a product.
3

Scale it.

Teams. Capital. Customers. Data. Platforms.

Unit 1,000 should be easier than unit 100.

Perspective

I've been on both sides.

Engineer Make it work.
Executive Build the team.
Founder Raise the money.
CEO Ship the product.

So when we talk, I'm not only asking whether something is technically possible. I'm asking:

Should you build it?

Can you afford it?

Can you ship it?

Will someone pay for it?

The Difference

Most diligence stops at the deck.
Mine starts at the architecture.

I'll look at the system. The BOM. The roadmap. The team. The technical moat. And the assumptions nobody has challenged yet.

That's where expensive mistakes hide.

Skim the deck

Trust the narrative

Debate the valuation

✓ Read the architecture

The Work

What I help with

Architecture

What should you build — and what shouldn't you?

Product

What's the minimum system that actually wins?

Hardware

Sensors, compute, power, networking, silicon, BOM.

Execution

Team, hiring, validation, manufacturing.

Fundraising

Turn technical complexity into a story investors understand.

Customers

Find out what they actually need — and will pay for.

Who I Work With

Founders building machines that sense, decide, and act.

Robotics Autonomy Industrial AI Automotive Defense Logistics Energy Infrastructure

Especially companies where hardware makes the software better — and the software makes the hardware more valuable.

Why Efficiently?

Because the best time to find a $10M mistake is when it costs $10K.