We first met Diego and Kenny in 2014 after being introduced by our friend Joe Heitzeberg, and immediately knew they were founders we wanted to work with. Their idea resonated. The world needed an app store for algorithms. As all applications become intelligent applications, better ways were (and are) needed for developers to quickly and efficiently tap into intelligent algorithms – especially for machine learning.
Kenny and Diego’s vision was to bring together the largest collection of algorithms in the world, each with their own product detail page and community reviews, each accessible with a consistent API, each modifiable and combinable, with production-ready performance and reliability, no friction to get started, and affordable to use deeply and broadly. And unlike most marketplaces where demand does not care until there is supply, and supply is not interested until there is demand – Algorithmia had a an elegant way to bootstrap their marketplace. “Supply” was sourced from open source, from universities, and from bright algorithm developers around the world who did not have an outlet for their work, much less monetize it. The Algorithmia marketplace launched in 2015. Fast-forward two short years and over 3500 algorithms are now in the marketplace with almost 50,000 developers using the system. The company’s early investments making Algorithmia one of the best platforms in the world to do ML and deep learnings have paid significant dividends.
Part of Kenny and Diego’s thesis has always been that enterprises and large organizations would want to tap into the Algorithmia “brain” of algorithms. They started receiving signal to validate this idea early on. Enterprises began to ask:
“Can we have our own Algorithmia?”
“Can we get an Algorithmia-in-a-box that we can populate with our own algorithms and also pull in yours? We have a difficult time sharing our in-house developed algorithms across divisions and groups, which creates a huge cost of re-inventing the wheel.”
“We need infrastructure and a microservices platform that can allow us to be agile with our machine learning models, and spin up these workloads as quickly, reliably and cost-effectively as Algorithmia.”
Thus, Codex was born, with In-Q-Tel becoming one of the early customers. Now many enterprises are using Codex as their in-house machine learning platform. The synergy between Codex and the public marketplace is essential – every enterprise customer or prospect has already tried, if not discovered, the company via Algorithmia marketplace. Algorithmia today is truly more than a marketplace for algorithms – it is a microservices platform for machine learning both in the cloud and on-premise.
Kenny and Diego had a compelling vision, and they are also great full stack entrepeneurs. Maniacally customer focused, tireless workers, very technical with sound business instincts. They also have hired a GREAT team that has gotten more done, consuming fewer resources, than almost any team we have worked. This has all translated into excellent execution.
Now, we are thrilled to have the new Google fund focused on AI and ML lead Algorithmia’s $10.5M Series A. I am particularly excited to work on the board with Anna Patterson, a visionary engineering leader who has led large scale machine learning work and teams across Google, within Android, core search, and her own startup. The broader Google team and world-class platform will be an incredible help to the company.
Algorithmia fits into Madrona’s goal to fund smart teams early in large, growing markets within key thematic areas. The shift to DevOps focused, cloud native enterprise computing and the rise of intelligent apps powered by AI and ML technologies are the megatrends driving Algorithmia and have been core themes for Madrona over the past several years.
We could not be more excited to continue to help Diego, Kenny and the team build and even more rapidly scale Algorithmia. We are honored to be part of the investor group that includes other returning investors Rakuten, Osage Partners, as well as excellent new investor Work-Bench.
Whether you are a developer of any level – Algorithmia users range from university students and hobbyists to senior enterprise architects and data scientists — or an enterprise that needs a (better) ML platform, I highly encourage you to go give Algorithmia a try!