How Valere Labs' CEO Leverages AI for Impact

During the course of a standard workday for Guy Pistone, you might find him listening to a soaring Hans Zimmer track on Spotify, talking to a colleague in Croatia on the phone, or sending a quick thumbs-up text to his wife.

Regardless of what he’s doing, Pistone always has one thing in the back of his mind: artificial intelligence (AI).

AI technology is driving an increasing number of projects in his business, and he’s finding ways to utilize it in his day-to-day tasks, too. He uses AI to turn data into hiring insights, summarize long articles, write code quickly, and revise contracts.

He even uses it to create positive impacts for causes and groups that are important to him.

‍

Exploring AI for real-world applications

Today, Pistone often uses ChatGPT in his work. But he’s been using AI for over five years—well before a proliferation of generative AI tools hit the market.

While studying at Clark University, he began exploring how to leverage AI and machine learning (ML) in mobile applications. As a longtime basketball player—who spent time playing professionally before moving into the business world—Pistone developed a way to incorporate AI into his beloved game.

The result was two different apps, Elete and Fitivity.

Elete used machine learning, augmented reality, and computer vision to gamify basketball training drills.

Fitivity, on the other hand, used machine learning to drive their app store optimization campaign by reverse engineering the App Store ranking algorithms as they relate to specific sports and fitness keywords. The result helped Pistone yield 15 million organic users on the Fitivity platform.

Pistone built both apps into successful businesses and ultimately sold them—putting him on the right path to lead Valere Labs.

Growing an international agency while helping others

In his role as CEO of Valere Labs, Pistone manages a team of over 130 people spread across five countries. Much of his workday focuses on business development, as well as figuring out how he can continue to satisfy growing client demand for AI applications.

“I spend a few hours every day meeting with leads from Upwork over Zoom in order to see if Valere Labs will be a good fit for them,” Pistone said. “Then, I meet with my heads of sales and service. We talk about the projects we’re about to close on and analyze the risk profile or discuss how we’ll move forward.”

He also makes sure to devote a portion of his time to projects that are meaningful to him, or that have a positive impact on the world at large.

“My aim is to use 50% of my developers’ time to solve real-world problems,” Pistone says.

While this drive to use AI for real-world change has manifested in a few ways, one of the most notable examples is his support for Alzheimer’s research.

The cause hits close to home—Pistone lost his father to the disease in 2020. By the end of the following year, he’d entered Valere Labs into partnership with Johns Hopkins University. Valere helped the university develop an app that researchers can use to gather better data on Alzheimer’s patients—and glean further insights as to how Alzheimer’s disease works.

Upskilling in the world of machine learning

In addition to keeping a focus on real-world impact, Pistone also devotes a portion of his time every day to exploring recent developments in AI. He’s particularly interested in how these developments can further improve Valere Labs’ work.

“GitHub recently came out with Copilot, which offers suggestions as you write code by using deep learning and natural language processing,” he said.

“We adopted Copilot in two offices as a test—and it’s made us at least 30% more productive when coding.”

It's this kind of exploration and discovery that Pistone recommends to all aspiring AI developers—along with more concentrated study in specific areas of machine learning.

In fact, it’s his number one tip for any developer interested in working with AI.

“Start with the latest tools and platforms that are coming onto the market so rapidly and spend at least 10% of your time understanding what they do and the value they provide,” he said. “That time investment can yield huge returns for projects or just making you more efficient at work.”

The type of tools a developer decides to learn depends on their interests and career goals.

“You don’t have to tackle machine learning from a very technical level. You really just have to learn how to prompt AI tools. Take ChatGPT, for example. It’s a prompting tool. If you can prompt it correctly to help you write or debug your code, you can get a lot of value out of it,” Pistone said. "You can go to YouTube and watch videos about AI prompting and gain a better understanding of how to use it to your advantage."

However, if someone does want to approach AI development from a more technical standpoint, Pistone recommends they start their journey by learning Python.

“Python is the language on which a lot of machine learning frameworks run,” he said. “If you learn Python back-end development, you’ll be able to do software development and AI. From there, you can narrow in on deep learning, computer vision, natural language processing, or another area of machine learning.”

Pistone anticipates that understanding how to work with AI tools will be an important part of software development careers going forward—but AI won’t fully replace humans.

“There’s always going to be a need for highly skilled software developers and project managers, because clients aren’t always sure what they need,” he said. “It’s really skilled project managers and developers who help them get to that point—and I’m confident that AI will never be able to do that.”

This type of human guidance is a big part of what Valere Labs provides to its clients, even when they’re leveraging AI tools for more efficient coding.

Turning raw data into business insights

“Clients will come to us and they’ll say that they want to build an AI to solve a specific need. But the challenge in AI is knowing where you get the data from,” Pistone said. “So before we can create the end solution, we need to get that data.”

He advises that clients should consider all the following when first exploring AI in the workplace:

• What data do you have available to you? Is it proprietary to your company?
• How is the data stored? Is it in a spreadsheet, a database, or another file?
• What format is the data in? Is it text, images, or something else?
• What are the points of friction you’re trying to solve with AI? Are they internal or external?

From there, AI developers and strategists—like those on staff at Valere Labs—can create solutions, and even highlight areas of additional value that clients might not have been expecting.

“For example, we had a client—a recruiter—through Upwork years ago. They’d done a great job of aggregating data from job review sites like Glassdoor,” Pistone said.

His team made an app to display all the client’s data in a nice, readable way—but they didn’t stop there.

“In talking to the client, we established that recruiters never really know when a worker will leave a company. It’s an embarrassment of riches that needs interpretation,” he said. “But we know the answer is in this data. So we used AI to look at a specific company’s feedback and create a sentiment score. A recruiter can look at the score and get a feel for if workers might be ready to leave that company.”

Taking on new client projects

Valere Labs can apply similar processes and insights to client projects in multiple industries. Pistone’s team has worked with startups, research hospitals, and Fortune 500 companies focused on business operations, healthcare, children’s entertainment, and more.  

If you’re interested in learning more about how Valere Labs can help your company turn its data into valuable, actionable insights and applications, visit their website.