He started at the place that invented the transistor. He took one company to an IPO. Then he built another from scratch - and kept building it for over two decades while the rest of Silicon Valley chased the next thing.
Naperville, Illinois, 1980s. A young engineer at Bell Laboratories is designing switching systems that cannot fail. Not should not fail - cannot fail. Fault tolerance, built into the architecture, for the phone networks that carry everything. That obsession with reliability never left Arthur Chang. It just found a bigger stage.
Fast-forward through a career that traced the full arc of the Silicon Valley tech boom: executive roles across storage, servers, and remote access in the 1990s. Then the CEO seat at SoloPoint Communications, a telecom equipment company he guided all the way to an IPO. Then Cradle Technologies, where he ran a multi-core multimedia DSP company. Two companies, two exits, two very different market cycles - and through it all, a question that kept nagging: why was business communication still so fragmented?
In 2001, the year Gmail didn't exist yet, YouTube was four years away, and most businesses ran their phone systems on-premise hardware the size of filing cabinets, Arthur Chang founded PanTerra Networks. The pitch was simple: one cloud platform for all of a company's communications - voice, messaging, video, contact center. What the industry would eventually call "Unified Communications as a Service," or UCaaS. Chang was building it before the acronym existed.
What he built next wasn't just a product. It was a company culture anchored in the belief that in cloud services, a mid-sized agile company can out-innovate a giant every time. "Agility and innovation, NOT size, matters when moving to the cloud," he wrote in a 2017 blog post that reads less like a marketing brief and more like a manifesto. He has been proving it for over two decades.
Today, PanTerra's platform - now branded as Streams.AI - powers unified communications for mid-market enterprises with a suite that includes voice over IP, team messaging, video conferencing, and an AI-powered contact center. The company hits $43.5 million in annual revenue with 140 employees and a 99.999% uptime guarantee backed by 30-second response times from live, US-based support engineers. In a world of chatbots and offshore ticketing queues, that's a product decision that speaks louder than any press release.
I don't use an alarm clock.
The pandemic gave Arthur Chang an unusual test case. When COVID hit, six of his ten adult children - all with Bay Area jobs - moved back home. Six remote workers, one house, and a CEO whose entire business proposition was that distributed work should feel seamless. "This period where we were comfortable with the way we were working, not challenging our norms," he reflected at the time. His own household became a live demonstration of both the opportunity and the friction.
What he observed in those months shaped PanTerra's product roadmap: the loss of informal knowledge transfer, the "water cooler" conversations that onboard new employees without anyone realizing it, the gap between technology that works and culture that coheres. His answer wasn't a whitepaper. It was features - supervisory tools built into Streams that competitors don't offer, designed specifically for the manager who needs visibility into a distributed team without feeling like a surveillance state.
Chang's approach to AI follows the same measured logic. In a landscape where every SaaS vendor is racing to bolt on large language models and announce "AI-powered" features before they work properly, he preaches what he calls a "crawl, walk, run" model. Build customer confidence first. Solve real problems with real AI solutions. "We're not focused on leapfrogging for the sake of it. We're focused on solving real problems with real AI-based solutions," he told the Plugged In Podcast.
The results are tangible. PanTerra launched Connect AI for smarter video conferencing and an AI-powered contact center. Then Luna AI, a conversational AI receptionist that handles inbound calls - not a simple IVR menu, but a natural-language system that can qualify callers, route intelligently, and integrate with CRM data in real time. The product earned industry recognition including TrustRadius Top Rated and Buyer's Choice awards in 2025 and Five9's Partner Excellence Award in 2024.
Through all of it, Chang has maintained a point of view that is increasingly rare in B2B SaaS: that the person answering your support call matters as much as the feature parity in your changelog. The 30-second live response guarantee isn't a tier upgrade. It's table stakes at PanTerra, the kind of commitment that only makes sense if you genuinely believe reliability is a feature - a conviction that goes back, perhaps, to designing systems that cannot fail at Bell Laboratories, all those years ago.
He has also been a consistent public voice in the UCaaS space. As an author on Entrepreneur.com and a regular contributor to PanTerra's own blog since 2016, Chang writes with the directness of someone who has watched multiple technology transitions up close. His 2016 blog series dissecting the weaknesses of Office 365 E5 was pointed, granular, and backed by engineering detail - not marketing counter-programming. In a space where most competitor analysis reads like it was written by a committee, his posts read like someone who built the product and wants you to know exactly why the other one is wrong.
Agility and innovation, NOT size, matters when moving to the cloud.
Arthur Chang - PanTerra Networks Blog, 2017Where the engineering foundation was laid - systems thinking, signals, the physics of information flow.
Berkeley's EECS program - where hardware meets software and both meet theory. The crucible of Silicon Valley's technical imagination.
Stanford GSB - the finishing school for Silicon Valley's operator class. Strategy, capital, and the art of building companies that last.
"We're not focused on leapfrogging for the sake of it. We're focused on solving real problems with real AI-based solutions."
"The generation just entering the workforce has no patience with their employer's technology if they can figure a hack that makes work easier."
"Agility and innovation, NOT size, matters when moving to the cloud."