The front page of Pattern Computer’s website.

A mysterious startup helmed by well-known tech prognosticator Mark Anderson is coming out of hiding today, with a goal of using what it says are new discoveries in computing to solve some of the world’s biggest problems. It’s a big claim, but Anderson says the Seattle-area company has the early results to back it up.

The company, Pattern Computer, has been in stealth mode for more than two years. It’s so secretive that the company has been operating under a codename, Coventry, and its website is a blank page that reads only “A Revolution in Computing is Coming — Pattern Computer.” Today it is unveiling its new system that Anderson promises will open up the next frontier of capabilities for technologies like artificial intelligence and machine learning.

Pattern Computer is not selling a computer, but rather a cloud service. The company has its own data center and machines, but to its customers and partners it will look like a service. What sets Pattern Computer apart, Anderson told GeekWire in an interview, is its ability to perform pattern discovery rather than pattern recognition. That’s a small linguistic difference that means a lot in the tech world.

Mark Anderson. (Pattern Computer Photo)

Pattern recognition is generally how computers operate today. People tell machines what to pull out of a data set, and computers do it. With pattern discovery, Anderson said machines are let loose to explore information and come back with findings that humans — or other computer systems — haven’t thought of before.

“Instead of being incremental and perhaps scientifically boring, what you are finding are things that are exhilarating,” Anderson said. “They’re new; they’re a bit frightening; they’re different; and they’re probably very high value because of their unique quality.”

Pattern Computer chose healthcare as the first big challenge to tackle. At an event in San Francisco Wednesday, Pattern Computer executives will be joined by heavyweights from the biotech world to share the work they have been doing together.

They include genomics pioneer Leroy Hood, founder the Institute for Systems Biology, which was acquired by Providence Health & Services, a Seattle-based healthcare group. Also speaking will be Ben Brown, department head of Molecular Ecosystems Biology at the Lawrence Berkeley National Laboratory, and Larry Smarr, the founding director of the California Institute for Telecommunications and Information Technology, a partnership between UC San Diego and UC Irvine.

Anderson said Pattern Computer has been focused on cancer and personalized medicine so far. He teased that the company has “been able, in a few short months, to make major discoveries in both those areas.”

The company chose “biomedicine” as its first challenge primarily because it wants to make an impact on the world. Additionally, Anderson thinks the field gives the company a chance to show its mettle.

“We picked biomedicine for several reasons: it’s hard, and we think by entering an extremely competitive environment and showing we can find things in data that’s publicly available that haven’t been found by others, we therefore can prove out the fact that we can re-find things that others have taken years to find, and we can also find brand new things no one else has found,” Anderson said.

Comparing Pattern Computer’s offerings to leading artificial intelligence solutions, Anderson said his system can jump between categories, unlike most AI that has to be narrowly focused. The company will look to tackle other complicated areas in addition to healthcare issues.

Anderson wouldn’t give away the details under the hood at Pattern Computer, but he did say his team built nearly everything from scratch, from the hardware to the entire software stack. Everything except for the chips running the product — which Anderson didn’t want to talk about — is something new.

Anderson assembled an all-star team to make the concept a reality. Ty Carlson, a 12-year Microsoft vet who worked in a variety of roles at the company before spending four years at Amazon is a senior manager on the Amazon Echo team, is the chief technology officer and a co-foundMichaer. Former Yahoo CFO Ken Goldman is reprising that role for Pattern Computer. James Reinders spent 27 years at Intel and is an expert on the Xeon products. Michael Riddle, the chief systems architect, is co-founder of software company Autodesk and a “ninja in the world of programming extremely difficult and visually compelling things,” according to Anderson.

In total, Pattern Computer has 26 employees and contractors and has raised a nearly $6 million while keeping quiet about its plans. Its headquarters are in Friday Harbor, Wash., where Anderson is based, and the company just opened up an office down the road from Microsoft in Redmond.

Some of the participants in the 2015 CTO Design Challenge that paved the way for Pattern Computer. (From left) Mark Anderson, Larry Smarr, Brad Holtz, Steve Coy, Ty Carlson, Mike Riddle, Ken Kreutz-Delgado, Tim Stonehocker, David Schoenberger, and Hugh Bradlow. (Photo by Kris Krüg)

Anderson is known for his Strategic News Service, a publication focused on the intersection of technology and economics. The company says its readership and membership base includes Bill Gates, Jeff Bezos, Paul Allen, Elon Musk and other tech luminaries.

Anderson’s news service claims to have correctly predicted the global financial collapse in 2007, the collapse of oil prices and the rise of digital assistants. He has a scientific background, and his work also includes modernizing technologies like wireless coverage around the world.

Anderson hosts a tech conference every year called Future in Review, and at the 2015 edition we got our first look at Pattern Computer. Many of the people working for and with Pattern Computer were there — Anderson, Carlson, Riddle, Smarr — developing the idea as part of a design challenge.

Anderson said Pattern Computer has the potential to change not just technology, but the entire scientific method. He thinks the concept of pattern discovery can lead scientists to ask better questions through their hypothesis, generating better outcomes.

“If you can use pattern discovery to create the question, you would end up with much greater scientific results,” Anderson said. “So this isn’t just a theory about something, it’s not just one more opinion, this goes deeply into the entire discussion of how we discover things in the world and how science works.”

Like what you're reading? Subscribe to GeekWire's free newsletters to catch every headline

Job Listings on GeekWork

Find more jobs on GeekWork. Employers, post a job here.