RelationalAI introduces new Rel capabilities for Snowflake AI Data Cloud
4 companies named across 2 categories, one of 17 articles referencing Snowflake. Previous coverage: Coder Named to Redpoint Ventures’ 2026 InfraRed 100 (May 2026).
Companies mentioned
Best suited for
- Seniority
- EVP / SVP / VP / AVP
- Job function
- Chief Information Security Officer
- Persona
- Data / AI Operations Leader
- Buyer role
- Decision Maker / Budget Holder
- Decision stage
- Which to buy
- Adoption curve
- Early Majority
- Technology maturity
- Operational Expansion
- Industry
- Information Technology / Software & Services / IT Services / Cloud Services
Our classification, not the publisher's statement. Best suited for, not only for.
RelationalAI said it introduced new capabilities for Rel, its agentic decision intelligence system, to run natively in Snowflake’s AI Data Cloud. The update was detailed at Snowflake Summit 26 and focuses on giving decision agents context, reasoning, and post-training for action across operational workflows.
RelationalAI linked the release to an enterprise gap between what AI can do in software development and what organizations get from it across the rest of the business. It said the capabilities include a Rel App, prescriptive and predictive reasoners, conversational decision intelligence in Snowflake CoWork, and “push-button” post-training.
The Rel App is described as capturing a shared, governed representation of how a business works, including concepts, relationships, and rules for decision-making. Domain experts can explore the model, follow connections, ask questions in natural language, and reason through decisions grounded in data inside Snowflake. The release also includes a library of coding agent skills that work across Snowflake CoCo, Claude Code, OpenAI Codex, and GitHub Copilot.
RelationalAI said the general availability of a prescriptive reasoner targets constrained optimization problems, while a predictive reasoner applies graph neural networks to enterprise data in Snowflake to forecast outcomes including demand, churn, and asset failure. It said the prescriptive and predictive reasoners run together in a single workflow without moving data off the platform, and that the suite supports complex multi-domain reasoning with LLM-based reasoning and domain-specific reasoners. The company also described a private preview of “push-button” post-training for specializing open-source LLMs against an enterprise’s data and semantic estate in Snowflake, and said RelationalAI served as a launch partner in the Open Semantic Interchange initiative to port semantic models into Snowflake via OSI. “Just like humans, agents have difficulty knowing how to make good decisions,” said Molham Aref, Founder and CEO of RelationalAI.
Press release, provided by Globe Newswire on behalf of RelationalAI. Read the original.