The Real-Time AI Data Race Is On
Machines are smart. As we now apply an increasing amount of Artificial Intelligence (AI) to our machines through increasingly sophisticated Machines Learning (ML) models, our machines are getting smarter all the time. But it’s not enough. We don’t just need more intelligence, we need it according to specific vectors.
Key among those vectors are AI at scale, AI that is validated, secured and bias-free (aka explainable AI) and AI engines that are capable of computation analysis in real-time.
The question we must now ask is: should AI specialists develop more real-time competencies, should real-time data streaming specialists work to innovate new tiers of AI, or should the responsibility fall to higher-level data platform specialists or the hyperscaler Cloud Services Providers (CSPs) themselves?
DataStax thinks this challenge is a data platform play, but then – it would, the company is an enterprise DataBase-as-a-Service (DBaaS) tools and technology specialist with a foundation in the Apache Cassandra open source database. Significantly, the company is now referring to itself as a real-time AI business, not least because it has this month entered into a definitive agreement to acquire Kaskada a machine learning company with dedicated competencies in managing, storing and accessing time-based data to train ML models.
Should we think it completely normal and natural for a cloud-centric database company to put so much effort into bringing AI and ML to the fore across its platform? After all, databases power the backbone of most of the technology services that we all use every day – so, ergo, giving them an additional smart automation shift makes lots of sense, doesn’t it?
Yes, but (and it’s a very big but) the problem is that most machine learning architectures never reach full production because they are too slow, expensive, and complex to support large-scale applications. This means there’s a natural (okay, you could say unnatural) disconnect between trying to build ML and trying to build it into enterprise-scale applications and services like databases.
To address this challenge, DataStax will now add Kaskada to its own cloud services, which includes its scalable – Astra DB DataBase-as-a-Service built on Apache Cassandra and [software application system] ‘event’ streaming with Astra Streaming. This, claims DataStax, will give organizations a single environment for delivering applications infused with real-time AI using an advanced ML/AI model – one that is used by firms including Netflix and Uber.
Both DataStax and Kaskada have a track record of contributing to open source communities such as Apache Cassandra, Apache Pulsar and Apache Beam.
