As organizations get better at managing and using a wider variety of data, the more they will adopt and make use of AI. IBM General Manager for Data and AI Rob Thomas has said organizations can’t have effective AI without sound IA (Information Architecture). And one of the pillars of any IA is data management.
In this new era of data, databases are no longer considered the traditional system of record or datastore. Today, expectations are higher. Databases must be smarter. They should understand what is being searched and find the most relevant information while using the most optimized way to locate the data. As organizations try to become more nimble, they want a faster and simpler way to do analytics across these hybrid data stores without the expensive and time consuming efforts to copy, replicate, transform or move data, that is, no more extract, transform, load (ETL). They also must self-manage and self-heal, which reduces maintenance overhead.
In other words, what’s needed is an AI-infused database.
That’s the vision for IBM Db2; to become the AI database that can help power today’s cognitive applications. IBM Hybrid Data Management (HDM) that is built on the Db2 Common SQL Engine provides a platform to manage all data types across all sources and destinations.
Because of the software’s high reliability, resiliency and security, among other things, thousands upon thousands of organizations rely on IBM Hybrid Data Management and Db2 — the leader in the online transaction processing (OLTP), online analytical processing (OLAP) and big data segments — to run mission-critical applications. With this new vision, Db2, the core component of the hybrid data management platform, will also enable customers to accelerate their AI application development while automating some data management.
We’re positioning IBM Db2 as the database of choice for AI application developers and data scientists. For example, starting today, IBM Db2 has drivers for popular data science languages and frameworks including Go, Ruby, Python, PHP, Java, Node.js, Sequelize, and Jupyter notebooks, to enable developers and data scientists to build AI applications with Db2 data for the first time. These drivers are available now at GitHub.
It is also now easier for Db2 administrators and data scientists alike to explore new data sets within Db2 thanks to a new tool called IBM Augmented Data Explorer. This tool combines natural language querying capabilities and faceted search. As a result, developers use a search-engine-like experience that combines writing queries with natural language querying and auto-completion, to explore data sets. Now developers can accelerate exploration of new data sets without any need to know SQL but still use its power. Moreover, even if the data is spread across diverse hybrid data stores, developers and data engineers can focus on their work using the sophisticated capabilities of IBM Data Virtualization without worrying about ETL.
They can also create cognitive applications within Db2 using IBM Watson Studio, and train models irrespective of whether the data is on-premises (Db2) or on the cloud (Db2 on Cloud) or both, for true hybrid, multicloud support.
Developers can get a fast start with code samples that are already baked into Jupyter notebook. Going forward, data scientists will be able to do complex modeling and visualization using Db2 Graph, a capability that will be released later this year.
In addition, AI application developers will no longer have to worry about tuning databases for performance. The next release of Db2, due this spring, will feature a machine-learning-based optimizer in addition to its proven cost optimizer for maximum performance. Db2 will also be able to automatically manage the resources and schedule the execution for its workload using Adaptive Workload Management.
In short, IBM Db2 is a database that is built for AI and powered by AI, making 2019 an exciting year for its users.