IBM Begins Release of New Watsonx.AI and Data Platform

Two components of IBM’s watsonx.ai and data platform are now available, and the third is on the way. IBM on Tuesday announced the release of watson.ai and watson.data. The company also confirmed that watsonx.governance is on pace to arrive in October.

Furthermore, IBM revealed that 150 enterprise clients – including Citibank, NASA, Samsung and Wimbledon – have participated in the company’s beta and tech previews.

Unveiled at the recent IBM Think conference in Orlando, Florida, chairman and CEO Arvind Krishna hailed watsonx as a “groundbreaking platform” for building, deploying and managing generative AI capabilities at scale.

IBM’s watsonx allows partners to train, tune and distribute models with generative AI and machine learning capabilities. Under development for three years, IBM designed watsonx to manage the life cycle of foundation models that are the basis of generative AI capabilities and for creating and tuning machine learning models.

WatsonX.ai is a studio with a user interface that enables developers to train, validate, tune and deploy traditional machine learning and generative AI models. WatsonX.data is a data store on a lakehouse architecture that IBM said is optimized for governed data and AI workloads. IBM has described watsonX.governance as a tool for building responsible, transparent and explainable AI workflows. According to IBM, watsonx.governance will also enable customers to direct, manage and monitor AI activities, map with regulatory requirements, and address ethical issues.

Customizing Foundation Models

According to Kareem Yusuf, senior VP of product management growth for IBM Software, partners can customize and deploy pre-built foundation models provided by IBM or use their own models. Also, IBM has partnered with Hugging Face, which offers tools for building, training and deploying machine learning models using open-source code.

IBM’s Kareem Yusuf

“The models are pre-trained to support a range of natural language processing (NLP) type tasks including question answering, content generation and summarization, text classification and extraction,” wrote Kareem Yusuf, senior VP of product management growth for IBM Software. “Future releases will provide access to a greater variety of IBM-trained proprietary foundation models for efficient domain and task specialization.”

Foundation models are trained with massive amounts of data that allow for generative AI capabilities with a broad set of raw data that can be applied to different tasks, such as natural language processing.

“Instead of one model built solely for one task, foundation models can be adapted across a wide variety of different scenarios, summarizing documents, generating stories, answering questions, writing code, solving math problems, synthesizing audio,” according to an explanation by Stanford University’s HAI.

Watsonx.AI Potential

An early adapter of IBM’s machine learning models and natural language processing technology is Citibank, which has implemented IBM Watson Discovery, IBM Cloud Pak for Data and IBM OpenPages with Watson to transform the daily work of the bank’s 2,500 internal auditors.

Marc Sabino, a Citibank managing director who is head of innovation and chief auditor for internal audit, shared his vision for using large language models in his organization on stage with Krishna during the Think conference keynote.

“With the launch of watsonx, it would be fantastic if we could look at ways that we can standardize code or look at the way the code is written, or connect the network,” he said. “If you think about the number of developers that we have within Citi, how do we connect that network to make sure that they’re operating in the most optimized and quality way?”