The People Who Hated Their Own Call Center Built the Software to Fix It

For six years, Alex Levin and Rebecca Greene helped run one of the busiest customer operations in America. At Handy and then Angi - the home-services giant that connects people to plumbers, cleaners, and handymen - they watched millions of customers try to get help. Most of the tools they had to work with pushed the same conclusion: keep it low-touch, keep it cheap, keep people off the phone. Customers noticed. They usually didn't call back happy.

Regal is the company they built out of that frustration. Founded in 2020 in New York, it sells software that lets businesses build, test, and manage voice AI agents - the automated callers and answerers that handle sales, support, and operations conversations. The company's own framing is blunt about the origin story: built for CX leaders, by CX leaders.

01 / What it doesAn operating system for customer conversations

Strip away the branding and Regal is a platform for putting AI on the phone - and, increasingly, on SMS, chat, and in-browser voice - so that a customer calling about a late delivery, a medical refill, or an insurance quote gets an answer instead of hold music. The agents run 24/7, handle both inbound and outbound conversations, and can qualify a lead, book an appointment, chase a payment, or resolve a support ticket without a human picking up.

What sets the pitch apart is less the talking robot and more everything wrapped around it. Regal ships the unglamorous machinery a contact center actually needs to trust automation in production: A/B testing between agent versions, real-time monitoring, listen-and-barge (a supervisor can drop into a live call), automated quality assurance, call recording, conversation intelligence, and reporting. It is the difference between a demo and a deployment.

02 / The problemWhy the customer call still breaks

The contact center is one of the last big corners of business software that most people experience as pain. Wait times, phone trees that loop, agents reading from scripts, information the company already had but somehow can't see. The founders' insight was that this isn't mostly a people problem - it's a tooling problem. Their teams were boxed into low-cost tactics because the software made high-touch service expensive.

Voice AI promised to change that math, and 2024 and 2025 filled with impressive demos. The catch, as anyone who has shipped one knows, is that a voice agent that dazzles in a boardroom can fall apart on call number four million. Regal's whole thesis is aimed at that gap: make the thing manageable, measurable, and coachable, and it survives contact with real customers.

03 / ProductsWhat you can actually build with it

At the center is a no-code AI Voice Agent Builder - the marketing tagline is "Build AI Agents. No Code. No Limits. 24/7 Results." Around it sit phone, SMS, and chat agents, a Unified Customer Profile that pulls a caller's history into one view, a Journey Builder that fires the right outreach on the right event, and Regal Copilot, which whispers guidance to human agents on live calls. The older DNA is still there too: a sales dialer and branded caller ID, from the company's earlier life as an event-driven outbound calling tool.

Crucially, Regal is designed to sit alongside a company's existing contact-center stack rather than rip it out. That additive posture is a go-to-market choice as much as a technical one - it lowers the bar to trying an AI agent on a slice of call volume before betting the whole operation on it.

04 / CustomersInsurance, healthcare, and the unsexy calls

Regal's customer list leans toward regulated, high-stakes, high-volume industries: AAA for roadside and membership, Kin Insurance, Ethos, and Allstate in insurance, Ro, K Health, eHealth, and Perry Health in care, plus fintech and education names like SoFi, Varsity Tutors, and Harvard through Podium Education. The company says it processes interactions for millions of consumers a month.

The most telling detail is where the value actually lands. The flashy use case is a friendly sales call. The quiet, profitable one is collections, retention, and reminders - the conversations nobody on staff wants to make, running reliably in the background. Voice AI's real killer app, on this evidence, is operational rather than cool.

05 / The competitionHow Regal is different

Regal sits in a crowded and well-funded race. There are AI-native challengers like Sierra, Decagon, Cresta, PolyAI, and Parloa, and incumbents like Five9, Genesys, NICE, and Twilio that everyone is trying to modernize or displace. Regal's differentiator is its origin: the founders ran the phones, so the product is built around operator instincts - manage the AI agent the way you'd manage a rep.

Dimension	Legacy CCaaS	Regal's angle
Built by	Telecom vendors	Contact center operators
Core unit	Human seats	AI agents you coach
Rollout	Rip & replace	Additive to your stack
Who builds agents	Engineering	CX teams, no code

06 / MoneyThe funding trail

Regal raised a small seed, then a $38.5M Series A led by Emergence Capital in September 2022 - back when the product was still about branded, event-driven phone calls. In October 2024, as it repositioned around autonomous AI phone agents (the old regal.io now redirects to regal.ai), it raised another $40M from Emergence, Homebrew, and Founder Collective. The business model is straightforward B2B SaaS: subscription plus usage tied to conversation volume and channels.

07 / CultureCustomers are royalty (yes, that's a value)

Regal is built in New York with an in-office rhythm of three days a week, and it runs on five stated values: customers are royalty, fast execution wins, growth mindset, data beats opinion, and enjoy the journey. The crown in the logo is a hand-drawn nod to the first one. It is a tidy bit of company mythology - the customer as monarch, the operators as staff who finally get to redesign the palace phone line.

08 / Where it fitsThe bet on the AI contact center

The contact center is a vast, slow-moving market that has resisted reinvention for decades. A handful of companies are now betting it gets rebuilt around AI agents this decade, and the interesting question is who owns the layer that matters. Regal's wager is that the winner won't be whoever has the most lifelike voice - it'll be whoever makes the agent manageable at scale, on real call volume, for teams that already know how to run a floor. If that's right, the boring tooling is the moat.

There is a neat symmetry to it all. Two people spent six years on the wrong side of bad contact-center software, then built the tool they wished they'd had - and pointed it straight at their former selves. Whether Regal becomes the default operating system for customer conversations is still an open bet. The origin story, at least, is honest about the problem.