The Emergence of the Algorithm as a Service Model
Algorithms are hot again and now we need a marketplace model for them! The algorithm as a service (AaaS) is triggering a lot of buzz and interest in the industry. Conceptually, AaaS provides a model for distributing, operating and monetizing algorithms that can be used from third party applications.
While the AaaS model was mostly seeing as a theoretic exercise, platforms like Algorithmia have helped to make it a reality for developers and algorithm authors. Just a few hours ago, Algorithmia announced a partnership with In-Q-Tel to provide a private algorithm sharing platform for the intelligence community. While Algorithmia is just one platform, its recent popularity have served as a strong validation of the AaaS space.
5 Factors Contributing to the Raise of AaaS
AaaS is far from being a new concept but is just recently that it has been achieved meaningful relevance within the cloud space. From a market perspective, there are several factors that are contributing to the momentum in AaaS technologies. Let’s explore a few of those factors:
We are Living in the Golden Era of Algorithms
Without a doubt, we are experience an renaissance in the algorithm space. Rapidly growing technology segments such as IoT, machine learning, mobile, big data are increasingly relying on algorithms to deliver a unique value proposition. As a result, we are seeing an explosion in the creation of new algorithms as well as a reduction of the times to productize algorithms.
The Growth of Machine Learning and Artificial Intelligence
Machine learning(ML) and artificial intelligence(AI) are the two main drivers behind the emergence of the AaaS model. While a few years ago, ML and AI stacks were based on a well-known set of algorithms, today we are constantly seeing new breakthroughs in AI or ML algorithms which are actively used in production-ready systems. If you visit an algorithm marketplace like Algorithmia, you will find that 90% of its algorithms are focused on ML and AI problems.
A Bridge Between Research and App Developers
The AaaS model represents a natural bridge between the algorithm authors and app developers. Traditionally, the process of productizing complex algorithms in software products could take years and, many times, the algorithms were hidden behind some specific product feature. The AaaS model offers native mechanism for authoring, publishing and operationalizing algorithms in a few hours without the need of investing in heavy infrastructure.
Algorithm Monetization
AaaS offers a monetization model for algorithms which simplifies the commercialization of these artifacts. Traditionally, there hasn’t been a clear path in academic or industry environments that allow researchers to monetize algorithms in an easy manner.
From Algorithms to APIs in No Time
Creating unique algorithms is hard but using them programmatically could also be painful. AaaS offers native mechanisms to expose algorithms via programmatic APIs that can be used from any language or platform. Additionally, AaaS enables the mechanisms for monitoring and controlling the runtime behavior of algorithms without requiring any additional infrastructure.
These are just some of the ideas I believe are contributing to the recent popularity of the AaaS model. One thing is for certain, the best times for AaaS are still ahead of us.
