Entrada Ventures is excited to have been in a position to lead the recent Series A round in MixMode, Inc. MixMode is an excellent example of all the things we are looking for in a company. The company has a strong team that we've known and worked with for years, innovative and exciting technology, and a market that is large and getting larger. And, of course, being right down the street here in Santa Barbara makes it convenient and easy for us to work together.
MixMode is led by a successful trio of entrepreneurs who've worked with and around the Entrada Team for the past 15 years. Starting with CEO John Keister, a serial entrepreneur with 20+ years of experience in which he previously co-founded two search and analytics companies that each grew to $100+ million in revenue and went public. John has also been a successful early-stage investor for the past 10-15 years and has led and/or made several angel investments with us before we founded Entrada Ventures. Chief Scientist and CTO Igor Mezic has spent his career developing highly complex algorithms and artificial intelligence for data analytics. Mezic has been working on the kinds of machine learning and artificial intelligence that are core to MixMode's technology for over 20 years, both with AIMdyn and at UCSB (primarily for DARPA and the DoD). Bryan Elliot, VP of Engineering, is an entrepreneur and technology architect with an emphasis on large-scale, distributed software systems, who has specialized in security (both software/internet security and physical security) and cloud computing. Previously he managed software teams (and teams of teams) as co-founder and CTO of Ping Identity (NYSE: PING), as well as Citrix-LogMeIn.
MixMode has developed the first cybersecurity platform based on "third wave" AI (as defined by DARPA: The Defense Advanced Research Projects Agency). The critical differentiator between third wave AI and other types of ML or AI is that third wave AI learns in an unsupervised manner (i.e. no rules, no training data, etc.). Unsupervised Learning works by drawing inferences from datasets without labels and is best used to find patterns when a system doesn't know what it's looking for. This facet makes it ideal for cybersecurity where threats and methods are continually changing, and current rules-based systems become outdated very quickly.
MixMode's AI runs in the background and learns which network behaviors are 'normal' and which are not, and surfaces critical warnings quickly, all without human intervention (unsupervised). In other words, this is not a rules-based security system requiring constant maintenance and updating, nor does it require a long (6-24 month) training period like most ML/AI systems on the market today. This means that after a relatively simple remote install, MixMode's AI can start learning about a customer's network traffic, nodes, and behaviors and quickly and autonomously begin protecting a network. MixMode's platform can significantly reduce alerts when compared to competitive systems, detect zero-day threats, and monitor cloud and multi-cloud systems allowing security teams to save 90-95% of the time they currently spend chasing false positive alerts and generate an immediate ROI while simultaneously reducing risk.
The result is that customers will start seeing improved security and reduced alerts in significantly less time and with less effort than with other solutions that are in the market today. This is critical in a time when more and more sensitive information is stored in multiple environments – and breaches are business and PR nightmares. It's no longer a case of "if a company will be attacked," but "when" and MixMode is in a position to help customers of any size and scale better detect and prevent cyber-attacks – even as the attackers evolve their methods and potentially incorporate AI themselves.
If you'd like to learn more about MixMode and its innovative solutions, here is a link to some case studies and whitepapers they've published: https://mixmode.ai/resources/