Founder Series: An Interview with Tom Doyle, CEO of Aspinity

The industry of voice-enabled devices is booming. Just look at all the devices that are integrated with Amazon Alexa, Google Assistant, and Siri.

These devices provide convenience in our day-to-day activities. It’s great to play music, make a call, or create a reminder without touching a single button.

The downside?

The battery life tanks. After all, the device is constantly listening for voice commands and digitizing everything it hears, even though most of that audio input isn’t relevant.

A team of engineers at Pittsburgh-based startup, Aspinity, set out to change this.

They developed a technology called RAMP which can distinguish a person’s voice before that sound data is digitized. The battery of an always-listening device can increase by up to 10x with this technology.

That’s huge for the industry of voice-first devices.

We talked with Aspinity CEO, Tom Doyle, to learn more about the technology, the problems they’re solving, and the kind of engineers they’re collaborating with to take their product to the next level.

The conversation below has been edited for length and content.

 

What product are you all building at Aspinity and what problems does it solve?

Over the last 30 years, there’s been improvement in the area of digital processing. But it’s run out of steam.

There needs to be a different approach when it comes to how we process mobile battery-operated products.

That’s where our team at Aspinity is going.

Our team is building Analog Machine Learning (ML) chips, which is a bit of a new concept. Many people understand analog. It’s a tough technology, but a lot of people are learning it. The other side is machine learning – the idea of being able to take information, classify it, and detect events and signatures.

We’ve merged those two.

Typically, machine learning algorithms run on large compute farms and digital processors.

But we are able to run those machine learning models on our analog based processing core. The reason analog is so important is because any voice detecting products (think Airpods, voice remotes, etc.) are picking up natural sound data, which is analog in nature.

If we’re able to stay in the analog domain, instead of taking the data and digitalizing it, then the data can be analyzed on low power hardware.

So, imagine earbuds lasting a much longer time, or TV remotes that save battery power by only waking up to your voice.

It’s a really powerful technology and is really game-changing in the sense that we’re moving this machine learning capability to a new paradigm in analog.

 

What are the skillsets that you’re looking for in engineers who could build that next level of growth at Aspinity? 

Our technology includes both analog and machine learning. So as we grow, we’re looking for people who can build layers of interface between the analog world and the software, machine learning world.

But those are two different disciplines.

On the one hand, there are a number of machine learning experts. But they don’t necessarily have a background on the analog side. There are also many engineers with an analog background who don’t have machine learning or software experience.

We really look for folks who have a background in one of those disciplines, but who also have enough experience in the other to propel them over the finish line. That enables them to be productive with what we are doing.

For engineers who have a discipline in one of those two fields, I recommend expanding your horizons because that really helps you propel to the other side.

 

What qualities does an engineer need to have to succeed in a young, early-stage startup?

When you come into a new, small company it’s very different than going to a big company. In that large corporate environment, you could end up doing just one thing for years.

The breadth of the tech stack is what I see as a great opportunity. Engineers aren’t stuck doing a single task; they’re exposed to a variety of technologies and languages.

At a company like ours, we’re very reactive to changes in the environment. That includes changes to what customers want and changes to what we’re focused on.

Because of that, many of our engineers are multi-disciplined. They wear multiple hats.

When I came into the startup space, I saw it as an opportunity. I liked that instead of always doing the same thing, I was equipped to do multiple different jobs.

Author: Lauren Alexander

More Related Insights