Domino Data Lab, the provider of the Enterprise AI platform trusted by over 20% of the Fortune 100, is launching new capabilities for Domino’s AI Project Hub, enabling enterprises to accelerate AI innovation and production with scalability and safety in mind. As part of the fall 2023 platform release, Domino reimagines the AI Project Hub by incorporating a robust ecosystem—which sees participation from cutting-edge AI companies—aimed at reducing time-to-value for AI projects and increasing overall AI responsibility.
The explosive success that AI has seen in the past year, particularly regarding generative AI (GenAI), has given rise to nearly every organization pushing for AI innovation. While “everyone can do AI proofs of concept,” getting these projects into production quickly and safely is another challenge, according to Nick Elprin, CEO and co-founder at Domino.
Domino’s latest release is driven by the company’s mission to “enable enterprises to unleash AI responsibly at the speed of business,” said Ramanan Balakrishnan, VP, product marketing, Domino. “The Fall 2023 update is all about delivering speed, scale, and governance in AI development, allowing our customers to create significant business impact responsibly.”
The platform update consists of these three AI acceleration components:
- Overcoming the “Where do I start?” hurdle, jumpstarting AI initiatives with prebuilt projects and industry best practices.
- Supercharging data science productivity via advanced automation and GenAI coding assistants.
- Simplifying data access and governance across the enterprise.
With pre-packed reference projects derived from the AI Project Hub’s new ecosystem of AI companies, organizations can access templated projects that are adaptable to their unique needs, IP, and data. From there, enterprises can build various AI applications—including those for text generation, enterprise Q&A chatbots, sentiment analysis, predictive modeling, and more—that are instilled with industry best practices for security from the start.
“The Domino platform is built to be secure by design and that naturally extends to sharing and reusing projects in our AI Project Hub,” explained Balakrishnan. “These projects adhere to industry best practices for security and governance and run in self-contained isolated environments to keep code and data secure.”
“In addition, we proactively test our software to find and fix security vulnerabilities before they can be exploited. Enterprises can share their projects securely with other teams and ensure standardization and best practices are enforced,” Balakrishnan continued.
Templates available for use from Domino’s AI Project Hub span a variety of use cases, including natural language processing, computer vision, generative AI, and more. All models are fully customizable and can be fine-tuned or trained with new data, as well as adjusted for “specific goals, data sources, models, hardware tiers, environments, prerequisites, and starter code,” said Balakrishnan. These templates are suitable for several advanced application projects, such as fault detection using computer vision and GenAI apps leveraging the latest foundation models from some of the most popular providers.
The AI Project Hub further emphasizes its community of industry leaders contributing to the hub’s wide breadth of models, which include:
- AWS for summary and email text generation using Amazon Titan, Anthropic foundation models, and LangChain
- Fiddler AI for evaluating LLM strength, accuracy, and safety
- ai for simplifying and accelerating AI apps for code, image, and text generation
- Artefact for generating structured, data-driven digital marketing reports
- KSM Technology Partners for accelerating BioPharma discovery and development
The AI Project Hub will continue to expand its ecosystem to cover many different sectors, software, tools, infrastructure, and foundational models, from both open source and commercial vendors.
The platform update is not only concerned with accelerating AI time-to-value and time-to-market but also the governance of the data it’s fed. Domino’s dedication to responsible and safe AI—which is critical for GenAI use cases—is manifested as immediate, governed access to the most prevalent data sources across the enterprise, both on-prem and in the cloud. These data governance and access functions include:
- Connectivity for more than a dozen sources, including Databricks clusters, IBM DB2 and Netezza, SAP HANA, and more
- Creatable data connections to external data platforms by admins using a service account for which only admins hold the credentials, which enables admins to grant access to the data source for users without needing access to their credentials
- Domino Data Sources set up by admins that require users to use their individual access credentials when that user connects to a Domino Data Source
- Domino Data Sources shareable across all users on a platform or a selected set for data source reuse
- Credentials stored securely in Domino
- Data Audit Logging for providing teams with complete visibility of all data sources accessed from development to production deployment
- Domino Model Sentry for transforming policies into responsible actions with governance across the entire data pipeline and holistic model monitoring across the model lifecycle
Domino’s extensive governance and access capabilities aim to eliminate data movement, transformation, and integration complexities, all while reducing time-to-insight with rapid data access and ensuring data responsibility with sensitive data.
Existing and prospective customers of Domino “should be thrilled about the unique balance of flexibility and agility data science teams need, combined with the control and governance the enterprise requires that Domino offers,” said Balakrishnan. “[The fall platform update] gives data science teams the tools to not just experiment but to scale those experiments into real-world applications.”
“It's about achieving rapid innovation without sacrificing responsibility or control,” Balakrishnan concluded.
If an enterprise is interested in becoming a partner, please reach out to aihub@dominodatalab.com. To learn more about Domino’s latest release, please visit https://domino.ai/.