The Rallie Story: A Tennis Ball Machine Built Out of Frustration

From Factory Floors to Founding Startups

Hi, I'm Sophie Luo, creator of Rallie. My journey into hardware started early—my dad is a mechanical engineer, and I grew up breaking and fixing things in his garage. As a kid, I'd walk factory floors with him during his overtime shifts, fascinated by how things were built. I always knew I wanted to create something of my own.

I went on to study Computer Science at UC Berkeley and the University of Pennsylvania, but my passion for building physical products never faded. My first real dive into hardware was at Rokid, a startup where I became the first product manager. We developed AR+AI glasses years before Meta's Ray-Ban glasses were even a thing, and today, Rokid is a unicorn.

Later, I founded RosePi Robotics, where I built AI-powered autonomous robots to improve efficiency in utility-scale solar farms. But despite our efforts, the market was tough, and we eventually ran out of cash. It was a painful but invaluable learning experience.

From Startup Failure to Tennis Obsession

After that, I took a break—and finally had time to dive into something I'd always wanted to pursue: tennis.

But I quickly ran into three big frustrations:

• 1️⃣Coaching was ridiculously expensive—$120+ per hour for private lessons or $50 for group sessions in the Bay Area.
• 2️⃣Finding hitting partners was hard—especially ones at my level.
• 3️⃣My schedule was too unpredictable—so it was tough to commit to regular practice.

Still, I loved the game. That winter, I was on the court every single day, taking lessons and training whenever I could. I quickly progressed from a beginner to a 2.5, but then I hit a wall. I didn't want to keep spending thousands on lessons, but I knew I needed repetitive practice to refine my strokes. That's when I started looking into tennis ball machines.

Why Do All Ball Machines… Suck?

I thought a ball machine would be the perfect solution—after all, even professional players use them for training. But when I started testing different models, I was shocked at how bad they were.

I bought one of the most popular brands and immediately ran into problems:

• ❌It only had one motor, so it could only serve topspin
• ❌It was way too heavy and too bulky, loading and unloading into my car and pulling it to the tennis court (1/2 mile walk) was a workout in itself
• ❌The shots were inconsistent—balls got stuck, and I never knew what to expect on the other side of the court

I then rented and borrowed models from Lobster, Spinfire, and even newer brands. They were even heavier, bulkier, and cost $1,500+—and still weren't great. I even looked into crowdfunded ball machines, only to find that many of them never actually shipped.

"That's when I thought: 'Why don't I build my own?'"

Bringing Rallie to Life

With years of experience in hardware, manufacturing, and shipping consumer products, combined with my passion for tennis, I knew I was in a unique position to build something better.

I started by tearing down my own machine, studying its mechanical structure and electrical controls, and reimagining what a ball machine should be.

What Makes Rallie Different?

• ✅Trio-Motor System – It can launch topspin, backspin, and even sidespin
• ✅80 MPH Top Speed – Train like a pro
• ✅Internal Oscillation – For unpredictable, realistic ball placement
• ✅30% Smaller & 30% Lighter – More portable than any mainstream ball machine

I was able to achieve all this with:

• 🔹Precision motor control (servo-based for accuracy)
• 🔹Higher-quality materials (no cheap parts that break easily)
• 🔹A smarter, more efficient design

What's Next?

I'm now waiting for custom parts to arrive so I can begin assembly and testing of the prototype. Over the next few months, I'll be:

• 🔹Iterating on the design through multiple rounds of prototyping
• 🔹Finalizing the EVT (Engineering Validation Testing) by Summer 2025
• 🔹Moving into DVT (Design Validation Testing) – optimizing for mass production
• 🔹Hosting live test events in SF & LA for real users to try the machine

If everything goes according to plan, we'll begin production in early 2026 and start shipping by March 2026. Having worked with top-tier OEMs like Foxconn, Flextronics, and Goertek, I know how to navigate mass production, and I'm confident in delivering Rallie on time.

Future Product Plans

While our initial focus is on creating the perfect tennis ball machine, we have ambitious plans to make it a real smart robot in the future:

• 🎯Knows where the player is on the court, and serves the ball accordingly
• 🎯Adapts difficulty level of drills automatically according to how you play
• 🎯Cloud-based drill sharing platform for coaches and players
• 🎯Real-time analytics and coaching (yes, the real AI coach)

How You Can Help & Get Involved

Thanks for making it this far! And thanks for your interest in my story and in rallie! I really hope to involve people who really love tennis and really care about the future of tennis training in my process of building my machine. I need your help with:

🚀 Let's build the future of tennis training together.