BREAKING Yang Cai has led VisualOn since 2003 - 22 years, one company ENCODING Optimizer cuts HEVC bitrates by more than 50% AWARD VisualOn Optimizer named Product of the Year at NAB 2024 PARTNERS Integrated with Nvidia NVENC and Intel GREEN Lower bitrate means lower energy in every stream BREAKING Yang Cai has led VisualOn since 2003 - 22 years, one company ENCODING Optimizer cuts HEVC bitrates by more than 50% AWARD VisualOn Optimizer named Product of the Year at NAB 2024 PARTNERS Integrated with Nvidia NVENC and Intel GREEN Lower bitrate means lower energy in every stream
Founder · Video Technology · San Jose

Yang Cai

The computer scientist who has quietly run VisualOn for more than two decades - and is now using AI to make streaming cheaper and greener.

President & CEO Yang Cai, President and CEO of VisualOn
Yang Cai, speaking on adaptive encoding at an industry interview.
2003
VisualOn Founded
50%+
HEVC Bitrate Cut
3
CS Degrees
Ph.D.
UT Austin

A Long Game in a Fast Industry

Most technology founders in Silicon Valley move on. They start a company, sell it or fold it, and chase the next idea. Yang Cai has done the opposite. He co-founded VisualOn in 2003, took the roles of president and CEO, and he is still there today. Twenty-two years in the same seat, watching video move from feature phones to 4K over-the-top platforms, is an unusual kind of persistence in an industry that reinvents itself every few years.

That persistence rests on a deep technical foundation. Cai is a computer scientist first. He earned a bachelor's degree in computer science and technology at Peking University in China, a master's at the University of Kansas, and a Ph.D. in computer science at the University of Texas at Austin. A decade of study across two continents before he shipped a commercial product - the kind of long runway that tends to produce engineers who understand a problem from the inside.

Twenty-two years. One company. The technology changed completely; the person building it did not.

From Codecs to Companies

Before VisualOn, Cai was senior vice president of engineering and a board member at MedioStream, where he designed one of the industry's first real-time software MPEG-2 encoders. That work mattered beyond a single product. Moving video compression off dedicated hardware and into software was part of a broader shift that made digital media flexible and affordable. Earlier roles at Tera Systems and Avant Corporation rounded out an engineering career built on the hard, unglamorous layer of computing.

When he founded VisualOn, the company started in the world of media players - the software development kits that let apps play video reliably across a chaotic landscape of mobile devices, browsers and operating systems. It is invisible work. When it goes right, viewers never think about it. VisualOn's SDKs ended up embedded in countless video sessions, handling the messy details of playback, adaptive streaming and content protection so that app makers did not have to.

Betting on Adaptive Encoding

The most recent chapter is about compression again, but with a new tool. In 2023, at the IBC broadcast trade show in Amsterdam, VisualOn unveiled Optimizer, an AI-enhanced Universal Content-Adaptive Encoding engine. The idea is straightforward to describe and hard to build: analyze video frame by frame and let the system dynamically adjust the encoder to hit a target quality with the smallest possible bitrate. It runs in a single transcoding pass, which keeps it fast enough for live use.

The numbers are the point. Optimizer reduces H.264 bitrates by more than 40 percent, HEVC by more than 50 percent, and AV1 by close to 40 percent, all while aiming to preserve or improve visual quality. Lower bitrate is not just a cost line. Less data means less storage, less bandwidth, and less energy burned moving pixels around the world. Cai has leaned into that angle, framing efficient encoding as a way to make streaming both cheaper and more environmentally friendly.

"Since its introduction at last year's IBC, Optimizer has uncovered numerous use cases for our AI-enhanced CAE solution."

The industry noticed. Optimizer was named Product of the Year at NAB 2024, the big broadcast show in Las Vegas. That same year, VisualOn announced it had integrated the engine with Nvidia's GPU-accelerated NVENC to run H.264, HEVC and AV1 at production speed. A separate collaboration with Intel put the technology to work on live streaming efficiency. Rather than trying to out-build the largest chip companies in the world, Cai plugged VisualOn's software into their hardware and made both more useful. In 2025, the company partnered with Agora to demonstrate AI-powered ultra-low latency sports streaming at NAB - stretching the technology toward the demanding case of live events.

Candor About a Hard Business

Cai is not given to founder theater. In an interview recorded at IBC, he put the state of the market plainly: generating profits in this environment is highly challenging. That is a striking thing for a CEO to say on camera, and it fits a pattern. He talks about video the way an engineer does - in terms of trade-offs, constraints and measurable results, not slogans. When he discusses AI, it is not as a magic word but as a practical way to shave frames and lower a bill.

Put together, the picture is of a builder who found a problem worth a career - making high-quality video move efficiently - and stayed with it through every generation of the technology. The codecs changed, from MPEG-2 to AV1. The devices changed, from feature phones to smart TVs. The business models changed, from packaged software to streaming and OTT. Through all of it, Yang Cai kept VisualOn pointed at the same quiet goal: better pictures, smaller files, and a company that is still standing to ship them.

Reporting drawn from public interviews, VisualOn press releases, conference listings and professional profiles. Figures and quotes are attributed to those sources.

By The Numbers

How Much Optimizer Trims

H.264 bitrate reduction40%+
HEVC bitrate reduction50%+
AV1 bitrate reduction~40%

Source: VisualOn / Intel technical materials on the Optimizer content-adaptive encoding engine. Lower bitrate also lowers storage, bandwidth and energy use.

"Generating profits in this environment is highly challenging."
- Yang Cai, CEO of VisualOn, in an IBC industry interview
Frequently Asked

Questions About Yang Cai

Who is Yang Cai?
Yang Cai is the co-founder, chairman, president and CEO of VisualOn, a San Jose video technology company he started in 2003. He holds a Ph.D. in computer science from the University of Texas at Austin.
What is VisualOn's Optimizer?
Optimizer is an AI-driven, single-pass content-adaptive encoding engine that analyzes video frame by frame to cut bitrates by roughly 40 to 50 percent while maintaining visual quality and lowering energy use.
What did Yang Cai do before VisualOn?
He was senior VP of engineering and a board member at MedioStream, where he designed one of the industry's first real-time software MPEG-2 encoders, and earlier worked at Tera Systems and Avant Corporation.
Where did Yang Cai study?
He earned a B.S. from Peking University, an M.S. from the University of Kansas, and a Ph.D. in computer science from the University of Texas at Austin.
What awards has VisualOn's technology won?
The Optimizer content-adaptive encoding engine was named Product of the Year at NAB 2024.
Share This Profile
in · LinkedIn X · Twitter f · Facebook Instagram