Hi Reader,

Welcome to Hal9 Weekly — insights from our founder Javier on turning AI ideas into real products.

Each week, we’ll share practical frameworks, founder stories, and product updates to help you build faster.

Javier's Insights:

The Simple Test: AI or Automation?

Here's a question I get constantly from founders: "Should I use AI for this, or just regular automation?"

The answer isn't what most people think. It's not about what's trendy or impressive. It comes down to two things: cost and quality.

But most founders waste weeks debating this instead of just building. So here's how to decide in under 5 minutes.

The Fast Way to Figure It Out:

If you're moving fast (and you should be), do one of these two things:

Option 1: Ask ChatGPT or Claude to draft it. See if the output is good enough. If it works, you have your answer.

Option 2: Ask your team or vendors for estimates. "How much to build this with AI? How much without it?" The cost comparison will make the decision obvious.

Once you do either of these, you'll know pretty quickly which path makes sense.

The Simple Rule:

Here's the framework I use to avoid overanalyzing every single task:

Does this task require human reasoning?

If NO → Use regular automation.

Example: Sending a monthly invoice reminder. There's no thinking involved. Should I send it this month? Yes. Should I send it next month? Yes. We just set up a Slack reminder that fires on the last day of every month. Done.

If YES → Use AI.

Example: Writing a blog post about what we shipped this week. Someone needs to figure out what happened, what matters, what's worth sharing. That requires judgment and reasoning. That's where AI makes sense.

The Bottom Line:

If the task is repetitive and doesn't need judgment, that's straight automation.

If it requires thinking, deciding, or adapting based on context, that's where AI belongs.

Your job as a founder isn't to pick perfectly. It's to pick fast, build it, and iterate faster.

Stop debating. Start building.

What tasks are you trying to figure out right now? Hit reply — I actually read these.

Customer Success Story: Limber Health

From Clinical Vision to FDA Approval with AI-Powered Motion Tracking

Limber Health saw a massive problem in physical therapy: patients couldn't stick with at-home exercises because there was no feedback, no accountability, and no way for doctors to track progress remotely.

They wanted to build AI-powered motion tracking that could evaluate exercises and functional tests through a smartphone camera — clinical-grade quality, but accessible from home.

The Challenge: Build an AI system sophisticated enough to track 3D motion, assess patient performance in real-time, and provide feedback during exercises — all while processing video locally to protect patient privacy. And it had to work on regular smartphones, not expensive specialized equipment.

Hal9 helped Limber Health:

🎯 Develop an AI runtime with pose estimation models that track patient motion in 3D space using standard smartphone cameras

🔒 Process all video locally on-device to protect patient privacy — only performance data goes to the server

📋 Document, test, and validate hundreds of different exercises and functional tests

🏥 Submit for FDA approval for one of the functional tests developed with Hal9

The result? Limber Health is now pursuing FDA approval and launching new AI-powered features for their iOS and Android apps, positioning themselves as leaders in the digital health space.

"We thoroughly enjoyed collaborating with Hal9 as our trusted AI partner, from initial design to successful FDA approval."​
— Chief Medical Officer @ Limber Health

Let’s Talk

Got an AI idea and want some help building it quickly?

Let's chat!

Talk soon,

—The Hal9 Team