“You have to assume every company will have access to the same LLMs and voices. The challenge, then, is to build a company that thrives despite this reality” Five Questions with RegalAI’s Alex Levin on Building In a Fast Moving Industry and More

Homebrew makes investments by consensus — it works because there’s just two of us. We’ve done it this way for two reasons — first, it works internally given our style of decision-making and respectful but loud debate. Second, it matters to us externally that founders know it’s always Homebrew making the investment — never a situation where one of us was excited and the other one didn’t block it. To that end, backing Regal, now a leader in AI Powered Calls for your businesses’ sales, support and operations, was quick to mutual agreement. Satya and I were excited by the vision and the cofounders’ previous work experience together. Since the world of AI is moving quite rapidly, I wanted to check in with Regal CEO Alex Levin to ask Five Questions.

Hunter Walk: In a recent Five Questions the person I was interviewing said to think about your career as a story you’re telling about yourself. What story does your career tell?

Alex Levin: I remember studying the philosophy and psychology around consciousness in college and thinking I might pursue academia. One day I realized that I could study my topic for my whole career and never get to the end of what was already known. And at that moment, it was crystal clear to me that I had to find a career where I could get to the forefront of what was known more quickly so I could help contribute instead of feeling like I was always behind.

After graduation, I dove headfirst into tech startups as it was clear they were investing the future every day. I knew that one day I wanted to be an executive, or even the top executive. And I understood that if I didn’t know how the sausage was made-if I didn’t know how to build technology-I’d never be eligible for those top roles. So early in my career, I prioritized being close to engineering teams, teaching myself how to build software, and learning how to be a product manager.

When I came across Marc Andreessen’s article, “ Why Software is Eating the World “ in 2011, it felt like a vindication of the direction I had chosen, but I was already down that path.

I’ve experienced both large and small companies, and I’ve found that I thrive in early-stage environments. It’s like how they describe the professional leagues of any sport: “things move faster”. Also, there’s more opportunity, and you really feel like if you’re not there one day the project doesn’t move forward. I love that sense of accountability.

I’ve been fortunate to take on progressively senior roles, mostly on the commercial side, and experience scale, including reaching $1.5 billion in revenue at Angi. And now, I’ve taken everything I’ve learned and put it to the test by starting a company myself.

Looking ahead, I hope my story is one where I continue to take on challenging opportunities, create something where nothing existed before, and build companies that truly change their markets. And, ideally, not just once-but many times.

HW: Similar to you, I started Homebrew with a former coworker. How did your relationship with Rebecca evolve and do you remember the ‘we’re going to do this!’ moment?

AL: While we were at Handy, Rebecca and I both had our first kids, and I distinctly remember talking about starting something together soon afterwards. We spent months having general conversations, but eventually, we sat down and did a “Founder Dating Quiz”. We went through a list of 20 to 30 key questions-things like our working styles, how big we wanted the company to be, what a good exit looked like, whether we wanted to raise money, what could go wrong, and what we were most worried about. After that, it really felt more real as we knew we were aligned.

It still took some time to land on Regal as the company we were going to start as we had two or three other ideas we seriously considered (which is a story for another time). But when we talked to potential Regal customers about our vision, and how we wanted to change the way companies engage with their customers, the response was overwhelming. People were pulling for it. And once we felt that “market fit”, we jumped in fast.

HW: Regal is an AI-driven company that was founded prior to the ChatGPT launch. How have the last few years accelerated the roadmap? How has it challenged the company?

AL: Coming from a background in product and marketing before Regal, Rebecca and I were used to great tools that could tap into data sources, personalize customer interactions and even apply machine learning-it wasn’t really AI back then-to do some pretty smart things. Meanwhile, in contact centers, teams were afraid to change anything in their software for fear it would all break, let alone try using data for personalization or machine learning. So we knew it would be a bit of an uphill battle to get contact centers to change.

From the very beginning in 2020 we had a vision of better business-customer interactions powered by customer data + ml/ai + software for teams to make changes easily. In 2020 the avant-garde AI was for assisting human agents, not autonomously handling tasks, but as we started building, we knew customer data and orchestration had to be at the core of our platform, hoping that one day, autonomous AI Agents could take over the customer conversation instead of a human.

For years, though, AI just wasn’t good enough. But we kept pushing on it because human agents cause issues because they have other motivations, weren’t always well-trained, could quit, or just have an off day. The public might assume humans are the gold standard, but anyone who has actually operated contact centers with human agents knows it’s actually really difficult, and there are plenty of challenges with human agents.

Then, about a year and a half ago, everything changed. LLMs finally reached a level where they could understand and generate language and make decisions similar to humans. And we got to a point that our AI agent demo completely opened our eyes as to what was possible. We started focusing on creating an “omni-agent” system, one where both AI and humans could operate seamlessly. Build the policies, scripts, orchestration, guardrails once, and deploy them to both human and AI agents.

A year and a half ago, not many companies were ready for AI. They worried about hallucinations, customer reception, and whether it would really work. But AI agents have advanced incredibly fast, and the platform we spent four years building to help human agents take the right action actually gave us the best-in-market platform to operate AI Agents and we now have a huge advantage over companies just starting to build AI agents from scratch. So we are seeing a seachange at consumer companies — everyone is evaluating AI Agents in 2025.

The best part of this shift is that it’s made our GTM simpler. To move human agents into Regal required ripping and replacing their current contact center software. Now when we sell Ai Agents, we roll out without touching their contact center software, accelerating the sales process and implementation.

HW: Content marketing dominates the venture industry these days, much of it about Artificial Intelligence. Since you’re down on the field with the actual companies buying these products, what’s something that we’re missing? Share an ‘earned insight’ that you have from the work at Regal?

AL: Rebecca and I are not naturally public people, so for years, we didn’t focus much on marketing. That was a mistake. With our newly built marketing team (who have strongly encouraged us to get on our game and step up), we are finally getting in front of more customers, and Rebecca and I are starting to embrace building in public.

One simple earned insight is about how to start testing AI Agents in production. Every company wants to start small. And most gravitate to small use cases (like out of hours) to test. I always advise against that as even if that succeeds, it won’t have a material impact on the company and you will need to start again with a larger use case. Instead, pick the largest use case for voice (like your main inbound or outbound calls) and test your AI Agent on 1% of calls to start. Then as it succeeds, scale up to have impact immediately.

HW: What’s your advice for founders who might be starting companies today in the general AI space — let’s focus on those who are using AI to solve problems for customers (applied AI) versus lower in the stack base models or deployment infra.

AL: Rebecca and I are founders and angel investors in the AI Voice space. You have to assume every company will have access to the same LLMs and voices. The challenge, then, is to build a company that thrives despite this reality. Said otherwise, build a “thick” workflow or application layer that will provide value no matter what the LLM and voices do.

So much of what Regal focuses on, and what we invest in as angels focuses on role-specific or industry specific workflow tools for generative AI to solve problems that LLMs alone cannot. For example, integrating AI into customer data systems, and enabling actions (like processing payments, updating CRMs, or sending SMS messages). Or ensuring AI stays within its intended scope, provides accurate information, and recognizes when it should say, “I don’t know.”

These challenges exist across industries and roles, and won’t be solved by LLMs alone.

Thanks Alex! To demo Regal, or learn more about their product, head to their website. And if you want to join a fast growing company that cares not just about what they’re building, but how they’re building it together as a team, well, Regal is hiring.

GET ALL MY POSTS VIA EMAIL (first and free)

Bluesky has the JUICE -> https://bsky.app/profile/hunterwalk.com

Originally published at https://hunterwalk.com on February 20, 2025.