By Maryam Ashoori, Director of Product Management, watsonx.ai, IBM and Heather Gentile, Director of Product Management, watsonx.governance, IBM
As momentum for AI in the enterprise continues to gain steam, the reality of moving these projects into production has proven more complicated – as business grapple with issues of trust, quality, scale, and associated costs. Ultimately, for businesses to gain value from their AI investments, they need a way to build and deploy these projects in a more efficient and cost-effective way – with transparency and trust at the foundation.
MLOps, short for Machine Learning Operations is a set of practices for companies to automate tasks and deploy models efficiently, ensuring everyone involved from data scientists to developers to engineers can collaborate effectively and monitor and improve models for better accuracy and performance. IBM’s solution for AI and MLOps is watsonx, an open AI and data platform that combines data, an AI studio and governance. We are pleased to announce that IBM has been named a Leader in the IDC MarketScape: Worldwide Machine Learning Operations Platforms 2024 Vendor Assessment (doc #US51573824, November 2024).
According to the IDC MarketScape, “the platform emphasizes governance, collaboration, and automation across the entire machine learning life cycle, including data ingestion, model development, registration, deployment, validation, monitoring, drift detection, and alerting.”
IDC MarketScape vendor analysis model is designed to provide an overview of the competitive fitness of ICT suppliers in a given market. The research methodology utilizes a rigorous scoring methodology based on both qualitative and quantitative criteria that results in a single graphical illustration of each vendor’s position within a given market. The Capabilities score measures vendor product, go-to-market and business execution in the short-term. The Strategy score measures alignment of vendor strategies with customer requirements in a 3-5-year timeframe. Vendor market share is represented by the size of the icons.
Source: “IDC MarketScape: Worldwide Machine Learning Operations Platforms 2024 Vendor Assessment” November 2024, IDC # US51573824.
Adopting an AI and data platform built for business
Businesses intent on utilizing AI effectively and responsibly need to develop a platform and governance strategy as early as possible.
Within watsonx, enterprises can build custom AI applications regardless of their stage of AI adoption and usage. As an open AI and data platform, watsonx incorporates a variety of data, AI and ML/Ops tools, foundation models, and governance to deploy trusted AI for use cases like content generation, deploying chatbots, building custom agents, customizing models, unlocking insights and coding more efficiently.
The IDC MarketScape report highlights noted, “IBM’s commitment to responsible AI is evidenced by its leadership in promoting open AI tools and practices and its participation in the AI Alliance, an international community of leading technology and research organizations collaborating to advance open, safe, and responsible AI.”
AI application development
IBM watsonx.ai is an enterprise-grade studio for AI builders that helps simplify and scale the development of AI applications with easy-to-use tools and system prompts for building, deploying, and monitoring AI applications. It provides AI developers and model builders a complete developer toolkit with APIs, SDKs, third-party frameworks, models, tools, and runtimes to streamline the application development process using high-quality data, with built in governance and guardrails.
IBM empowers developers across all skill levels to get started quickly and build, deploy, and manage AI with a variety of pro-code, low-code, and no-code interfaces. Watsonx.ai offers capabilities like AutoAI to analyze data and automatically recommend models. The IDC MarketScape report noted, “IBM watsonx.ai empowers users with various skill levels to leverage AI/ML lowering the barrier to entry for tasks such as data preparation, synthetic data generation, and automated model building without the need for extensive coding knowledge.” Other productivity tools including watsonx Code Assistant compliment watsonx.ai, offering generative AI capabilities to optimize the software development lifecycle.
Within watsonx.ai, AI builders can access and customize a collection of foundation models, including IBM’s flagship Granite models, using tools like InstructLab to fine-tune LLMs with enterprise data, a Prompt Lab interface, a data science toolset to build ML models with code or AutoAI development, as well as access a collection of powerful visual data pipelines and flows, and synthetic data generation.
The progression of generative AI technology can make it difficult for developers to keep up and adopt the most recent and valuable tools. For developers looking to get started with watsonx, IBM recently launched the watsonx Developer Hub – a curated set of tools with various resources, quick starts and guides for use of our APIs and SDKs, so developers can quickly go from signing up to achieving their first use case.
AI Governance
Watsonx.governance compliments watsonx.ai by addressing compliance, risk and security management and AI lifecycle governance. The solution is platform agnostic, enabling it to govern any foundation model or traditional ML model, including both IBM and 3rd party models and platforms, regardless of where it is deployed.
Watsonx.governance provides automation of model monitoring across the AI lifecycle using AI-optimized tools to drive scalability, optimize costs and increase return on investment for all AI initiatives. The IDC MarketScape highlights our strengths in end-to-end lifecycle governance stating “IBM provides a centralized registry for model inventory and automatically documents key metadata. watsonx can continuously monitor AI models for quality, performance, fairness, and transparency even when models are built and deployed with non-IBM technology “
When it comes to risk management and compliance, users can establish tolerances and alerts for detecting risks such as accuracy, performance, bias, and drift and access automated GRC tools to cover operational risk, policy and compliance management. Users can simplify compliance with tools that automate the process of identifying changes to industry standards and regulations and translating them into enforceable policies, map regulatory requirements to internal risk data in a single repository, and access factsheets for a holistic view of the model lifecycle metadata to support regulatory audits.
watsonx.governance creates a complete audit trail, beginning at the time the use case is requested by the business, capturing governance information throughout the use case review and approval process including risk assessments and compliance assessments through build and monitoring at run-time. The IDC MarketScape noted for IBM, “Model risk management and compliance are also supported through various assessments and workflows including model development approval, model attestation workflows, and decommissioning workflows.”
AI-Ready Data
AI would not be possible without access to the right data, both structured and unstructured, integrated into the MLOps lifecycle. As part of the watsonx platform, IBM offers watsonx.data, an open, hybrid data lakehouse to power AI and analytics with all your data, anywhere. To protect data across the AI lifecycle, IBM offers robust data security capabilities to secure critical data used for AI from both current and emerging risks across the hybrid cloud. The IDC MarketScape noted “watsonx provides multilevel security mechanisms including network security, enterprise security, account security, and data security for protection of data, applications, identity, and resources.”