If you follow this blog, you know I am a big fan of Algorithmia. The Seattle-based company recently announced a new $10.5 million series A to expand its algorithm as a service(AgaaS) model into the enterprise. The round was led by Google’s new artificial intelligence(AI) fund as well as existing investors Work-Bench, Madrona Venture Group, Rakuten Ventures and Osage University Partners.
Algorithmia has pioneered the vision of AgaaS as an essential element of modern AI architectures. The closest analogy to AgaaS is the mobile app store concept which have served as the distribution backbone of the mobile app ecosystem. Similarly, agaaS provides a model to distribute and provision AI algorithms.
The best known capability of the AIgorithmia platform is its AI marketplace that contains about 3500 AI algorithms as well as complementarily functions for the implementation of AI solutions. Additionally, Algorithmia provides a private cloud version of its CODEX platform that can run AI models at scale. Feature-by-feature, Algorithmia can be considered one of the most creative offerings in the new AI technology ecosystem.
5 Reasons Why Algorithmia Matters
There are several reasons why Algorithmia represents a unique but incredibly relevant offering in the AI market in general and specifically for enterprises. Here are some of my favorite (some of which I listed in my previous article about Algorithmia).
1 — A Better AI Research Distribution Channel
AI researchers and academics are constantly creating new and innovative algorithms. However, a significant percentage of AI research foes unnoticed beyond small academic circles it is simply buried in the pages of AI scientific journals. Algorithmia’s marketplace offers a more practical distribution channel fro AI models that can be made available to mainstream users.
2 — Connecting AI Researchers and Enterprises
Indirectly, Algorithmia is providing a bridge for enterprise to have access to the latest AI research and explore new algorithms that can be applied to their specific scenarios. Algorithmia allow enterprises to effectively discover new AI models that can be used on their machine intelligence solutions.
3 — Practical Validation of AI Models
The pages of AI journals are filled with claims about the unmatched virtuosity of specific algorithms. Despite its theoretical contributions, many of those published AI models have never been validated in real world scenarios using industry data sources. Algorithmia provides an efficient platform to validate and improve the efficiency of AI algorithms in the real world.
4 — Peer Reviews and Contributions
Peer reviews are an important aspect of academic research and AI is not the exception. However, even the most prestigious AI publications constraint peer reviews to a handful of experts. Algorithmia collaboration features allow AI researchers and developers to regularly review and contribute to AI algorithms.
5 — Enterprise AI runtime
This blog is like a broken record when comes to preaching the importance of scalable runtimes for enterprise AI solutions. Algorithmia’s CODEX provides a runtime engine that can execute AI models on both public and private cloud infrastructures which is vital for the mainstream adoption of AI in the enterprise. Algorithmia joins the efforts of other innovative platforms such as Bonsai, H2O.ai or bitfusion in this area.
Is Google a Potential Acquirer?
The fact that Google led Algorithmia’s series A could be taken as a sign of a potential acquisition in the near future. In my opinion, Algorithmia’s AgaaS model is still new and needs to be validated. However, the Seattle-based company already has tangible assets such as over 45,000 developers and 3500 algorithms as well as a solid technology stack. Under Diane Green’s leadership, Google has shown to be an aggressive acquirer in the AI space and Algorithmia could be a very unique addition to Google Cloud and a strong differentiator against competitors from Amazon and Microsoft.