Supply: Shutterstock AI-Generated
For all of the advances in analytics, two issues proceed to plague information groups: misaligned or inconsistent metrics throughout instruments and the inaccessibility of unstructured information. These two points not solely decelerate decision-making but additionally power groups to depend on handbook workarounds – introducing inefficiencies, inflicting information drift, and rising the chance of pricey errors.
When groups outline KPIs in numerous instruments, reminiscent of dashboards, spreadsheets, and stories, these numbers usually don’t match, undermining belief within the information. In the meantime, vital enterprise data stays locked inside PDFs, contracts, and different unstructured recordsdata that almost all BI platforms can’t entry or analyze successfully.
Sigma, a warehouse-native BI and analytics platform, introduced two main updates on the Snowflake Summit 2025 that straight handle these ache factors. In partnership with Snowflake, the corporate unveiled native help for Snowflake Semantic Views and integration with AI SQL, Snowflake’s breakthrough function for querying unstructured information utilizing giant language fashions (LLMs).
These advances are designed to help ruled semantic exploration and allow AI-powered evaluation of unstructured recordsdata, straight inside Sigma’s spreadsheet-style interface. Sigma positions this as a big step ahead in unified analytics because it permits customers to question each structured metrics and uncooked human context side-by-side inside a single ruled system.
Based on Mike Palmer, CEO of Sigma, the combination with Snowflake Semantic Views is “really native” and is constructed for flexibility, scale, and the following technology of analytics. “By assembly the semantic layer the place it belongs, we’re giving enterprise groups on the spot entry to ruled metrics and logic with out compromise,” explains Palmer. “And that is just the start. From bi-directional syncs to visible semantic exploration, Sigma is constructing towards a unified modeling expertise that brings readability and management to each layer of the information stack.”
“This can be a main leap ahead in delivering a constant, ruled expertise powered completely by Snowflake,” added Palmer. “Sigma is constructing towards a future the place each layer of the information stack speaks the identical language – outlined as soon as, executed in every single place.”
The corporate presents this integration as a option to streamline analytics by aligning enterprise metrics straight with the information warehouse. It’s a part of a broader push to make sure that everybody, from analysts to enterprise customers, is working from the identical trusted definitions.
The collaboration with Snowflake additionally displays a broader business shift towards warehouse-native analytics. For Sigma, it’s an opportunity to assist understand a long-standing aim within the information world: outline your online business logic as soon as, maintain it ruled on the supply, and make it accessible wherever it’s wanted.
This implies no extra duplicating definitions throughout dashboards and spreadsheets. As an alternative, every thing stays anchored within the warehouse, giving groups a single, trusted basis to work from. That’s one much less supply of confusion when groups want to maneuver rapidly and belief their information.
The opposite main replace is Sigma’s help for Snowflake AI SQL, a brand new set of LLM-powered instruments that make it potential to question unstructured content material, like photos, scanned receipts, product manuals, or contracts, similar to how you’d in the event that they have been rows in a desk.
Supply: Sigmacomputing.com
This improve pairs properly with Sigma’s not too long ago launched File Column Kind, which permits customers to add and work with recordsdata straight within the platform. Paired with AI SQL, this opens the door to make use of instances that have been beforehand cumbersome, like extracting cost phrases from contracts or reviewing receipts as a part of claims workflows. The brand new capabilities scale back the necessity for customized scripts, third-party instruments, or handbook information entry.
“For many years, legacy BI instruments assumed your information was clear, structured, and ready politely in rows and columns,” stated Palmer. “However a number of the most essential enterprise selections are made with the messy stuff: authorized paperwork, compliance PDFs, screenshots, receipts, product specs, and annotated photos.
“Traditionally, these codecs required a human within the loop: to learn, interpret, and manually extract insights. That’s the bottleneck AI SQL removes. Sigma and Snowflake flip human data into scalable methods, unlocking completely new varieties of evaluation throughout industries and groups.”
For Snowflake, the combination helps advance its imaginative and prescient of a unified, AI-powered information cloud. It validates its semantic layer technique and showcases how third-party platforms can faucet into Cortex AI SQL, with out heavy engineering.
Carl Perry, Head of Analytics at Snowflake, stated the combination “marks an essential step ahead in enabling enterprises to leverage the state-of-the-art AI options accessible with Snowflake Intelligence and Cortex Analyst,” including that it helps groups “create extra environment friendly and highly effective workflows.”
The newest updates to the Sigma platform aren’t nearly stacking on extra options – it presents a glimpse into what fashionable analytics may be. With native help for Semantic Views and AI SQL, Sigma is displaying how structured metrics and unstructured context can coexist in a single interface.
Associated Gadgets
Why a Common Semantic Layer is the Key to Unlock Worth from Your Information
A Semantic Strategy to Huge Information Governance
AtScale Broadcasts Main Improve To Its Semantic Layer Platform