An engineering company had a small file and a large wait. At C&S Companies, opening a 1.5-megabyte CAD file over a wide-area network could take twenty minutes. After installing Panzura, the same operation took eight seconds. Randy Chou used this customer example in 2015 to explain what his company was trying to fix. A drawing smaller than many holiday photographs had become an appointment.
For Chou, Panzura’s co-founder and then CEO, storage was entangled with the way people worked together. An office could possess the right application, the right employee and the right drawing, yet still spend its morning waiting for the three to meet. His proposed answer was a shared file system that made geographically scattered work feel local.
Same file. Different morning.
The example gives his career a useful point of entry. Chou has worked on security, networking, cloud storage and decryption. These are subjects that seldom demand a place at the dinner table. Their consequences do. A stalled project, an inaccessible file or an encrypted system can interrupt everyone else’s plans. His work sits in that uncomfortable space between having information and being able to use it.
A computer with the lid off
Chou was born in Taiwan and grew up in Colombia, Bolivia and Spain. His father was a diplomat. Around twelve, he got a Commodore 64; when it broke, he learned by opening it and trying to repair it. Software and hardware arrived together. The machine was both something to program and something to understand.
His formal education included computer science at the University of California, Berkeley. His early professional record includes work as a 3D software architect at Broderbund from 1996 to 1999, followed by SSL software leadership at Alteon WebSystems. Those roles put graphics, security and networking into the same career before Panzura existed.
At Alteon, Chou led development of SSL security and load-balancing software. He subsequently joined Aruba’s founding software team and managed its security software work. Alteon was acquired by Nortel in a transaction valued at $7.8 billion; Aruba went public in 2007 and was acquired by HP in 2015 for roughly $3 billion. These were company transactions, with many people behind them. Chou’s contribution was in the engineering.
There is a more intimate trace of that contribution in an earlier invention. A wireless virtual private network patent application filed in 2005 names Chou alongside John Taylor and Pradeep Iyer. The work concerns encrypting and transporting information between remote systems. Taylor would later become his Panzura co-founder. Their working relationship had already reached the stage where the paperwork was being filed.
Making the cloud work with familiar software
Chou and Taylor founded Panzura in July 2008. By September, Matrix Partners had supplied $6 million in financing. An early company profile described Chou’s security background and Taylor’s engineering experience; the pair were taking expertise developed in networking into a young cloud-storage business. The timing offered no shortage of reasons to worry, but the company had a concrete technical problem to pursue.
Panzura emerged with an Application Cloud Controller in 2010. Its design included solid-state storage and an awareness of applications. The controller sat between software people already used and storage that might be far away. Keeping frequently needed data close to users was part of the effort to reduce delay. Chou’s pitch emphasized that applications did not have to be changed to work with it.
That last condition matters. Enterprise customers do not arrive carrying a blank sheet of paper. They arrive with established software, budgets, habits and a considerable investment in yesterday’s decisions. A new storage system has to negotiate with all of them. Panzura’s proposition gave the cloud a familiar front door rather than asking every user to learn a new way of working.
In 2012, Panzura raised another $15 million to support international expansion and wider enterprise adoption. Its PZOS3 solution won the Data Center and Storage category in that year’s Best of Interop awards. The funding and the award marked different kinds of recognition: capital to expand the company and attention for the product its engineers had built.
The promise was tested in industries that handle drawings, models and other awkwardly large files. Panzura’s global file system used locking to prevent one person from overwriting another’s work. A colleague could receive read-only access while someone elsewhere held the lock. Shared storage needs rules of possession, even when the participants cannot see one another. A digital drawing can have a remarkably territorial social life.
Success could look like nothing happening
Chou described an especially appealing outcome in 2014: users could work with Panzura without knowing they were using it. The system replaced the local file server while keeping a familiar Windows-style experience. Data infrastructure had done its job when the person opening the file had no reason to think about data infrastructure.
“Many of our end users don’t know they are using Panzura.”Randy Chou · 2014
This is an unusual standard for a founder’s achievement. Many products announce their arrival through a new screen or a new ritual. Chou’s storage work sought to preserve an existing ritual while removing the delay behind it. The user still opened a file. The engineering was meant to change what happened between the click and the result.
He also had a sporting vocabulary for persistence. Soccer interested him, especially the final minutes when the match was still undecided. He described staying in the contest a little longer as a route to a larger reward. It is a practical metaphor for the work of an infrastructure company, where a promising demonstration is followed by the less glamorous business of making the system dependable.
By 2013, Panzura’s financing had reached about $58 million, with backing from firms including Meritech, Matrix, Khosla and Opus, alongside Chevron Technology Ventures. Its international expansion plans included Asia Pacific after an earlier move into Europe, the Middle East and Africa. The company’s geography was expanding along with the geography of the teams it wanted to connect.
The founder leaves the chair
In January 2016, Chou’s tenure as Panzura CEO ended. John Peters took over on an interim basis. The board wanted new leadership for the company’s next stage, and Peters publicly credited Chou with getting it off the ground with solid technology. The transition put a boundary around a chapter that had begun with two engineers and a storage problem.
Patrick Harr became CEO later that year. He described collaboration as a strong base from which the company needed to expand, and brought experience from VMware and Hewlett-Packard. Panzura’s next phase belonged to a different executive. Chou’s role as co-founder remained part of its history.
The transition also separated two responsibilities Chou had held together: founding a business and running it. Panzura continued with a new executive in charge. Chou returned to security with another venture. His earlier experience traveled with him, even as the company he had co-founded moved into a new period under different management.
From sharing files to finding keys
Chou and Greig Bannister launched Nubeva in 2016 to address a gap in public-cloud monitoring. Organizations moving applications to services such as AWS, Microsoft Azure and Google Cloud still needed visibility into their network traffic. Nubeva’s early Prisms software handled packet traffic so security tools could receive it. The storage founder was back in territory shaped by his networking and security experience.
On April 16, 2018, Chou joined a market-opening ceremony in Toronto. Nubeva Technologies had begun trading on the TSX Venture Exchange on March 2. The photograph records a rare moment when a career spent working behind familiar applications became a public occasion: colleagues gathered around the exchange podium, with the company’s name on the screen.
Nubeva introduced Session Key Intercept in 2019. The technology supplied encryption keys for authorized inspection of encrypted traffic. In 2021, the company appeared in the TSX Venture 50 ranking. Its direction was becoming more specific: helping other security systems see data that encryption would otherwise conceal.
In September 2021, Nubeva announced that initial testing had validated an extension of its key-discovery work to ransomware recovery. The premise was to retain encryption keys so they could later be used to restore locked files. Successful recovery depended on obtaining those keys and applying them correctly. Encryption itself had not become weaker; the company was working on access to the means of reversing it.
Steve Perkins, who worked with Chou and Bannister at Aruba before joining Nubeva, described their interest in problems at the intersection of security, cloud and communications. The connection matters because this was a new company assembled partly from an existing professional network. Different products, familiar collaborators, and another effort to make a difficult technical operation useful to customers.
The next problem is already in pieces
In March 2024, Nubeva announced a binding agreement to sell its TLS assets for $1 million and concentrate on ransomware reversal, with AI supporting recovery and automation. Bannister was to step down as CTO and board member in connection with the sale. Chou continued as CEO. The company was narrowing its work rather than carrying every earlier product forward.
A patent issued on September 30, 2025 named Chou as inventor of a ransomware-recovery method. Another followed on February 24, 2026: U.S. Patent 12,561,431, assigned to Nubeva, for recovery using computational intelligence. The latter addresses storage encrypted in multiple subsections, with harvested keys and pattern analysis used to recover those sections. A large file can become several smaller puzzles.
U.S. patent issued February 24, 2026
Assigned to Nubeva, Inc.
The patent places Chou’s name on the technical work as well as the management chart. It also brings storage and security back into the same frame. Earlier, he worked to keep shared files accessible across distance. Here, the problem is recovering access when portions of stored data have been encrypted. The interruption is different; the need for usable information persists.
On September 3, 2026, Nubeva announced access to a frontier AI model for defensive cybersecurity evaluation. The company said the work was at an early stage and did not promise a resulting product or revenue. For Chou’s next chapter, the useful detail is the experiment itself: applying another tool to the stubborn mechanics of detection and decryption.
His career can be followed through company names and transaction dates. It can also be followed through the everyday operations his teams have tried to improve: opening, sharing, inspecting, recovering. Those verbs explain why the engineering matters. A file may be tiny, an office may be far away, an encryption key may be hard to find. Someone still needs to get back to work.