Sigma Computing vs Holistics
Side-by-side trajectory, velocity, and editorial themes.
Sigma builds out the agentic analytics stack: workflow automation, Snowflake Cortex bindings, and a push beyond read-only dashboards.
Sigma is leaning hard into agentic analytics positioning. Recent shipments — Automated Actions for scheduled workflows, Sigma Skills accessible inside Snowflake Cortex Code, and bidirectional JavaScript events for embedded analytics — combine into a story about analytics that act and integrate, not just visualize. Concurrent thought-leadership pieces reinforce the messaging that read-only dashboards are insufficient for modern enterprise AI.
The platform is converging analytics, AI agents, and Snowflake-native tooling into a single operating layer. Investments are flowing toward workflows that trigger actions on schedule (and likely on events next), tighter Cortex integration so data engineers stay inside Snowflake, and embedded analytics primitives that let host apps surface and react to in-Sigma activity. The Gartner agentic AI mention is being amplified to support sales positioning into 2026 enterprise budgets.
Expect Sigma to add event-driven triggers and broader agent tool-calling to Automated Actions, and to deepen the Cortex bridge so a Snowflake developer can author and govern Sigma workbooks/data models without leaving the warehouse environment.
Holistics turns the BI dashboard into a conversational AI surface, on customer-owned models.
Holistics is well into a BI-meets-AI productization phase, layering conversational analytics on top of its existing modeling and dashboard core. Recent releases mix consumer-grade dashboard polish (auto-run filters, K/M/B number formatting, percentile calculations) with deeper AI plumbing: bring-your-own Claude and Gemini keys, per-user AI access controls, and now an Ask AI that asks clarifying questions back. The GitHub App integration also signals enterprise-readiness work alongside the AI push.
The product is being repositioned from a self-service BI tool to an AI-mediated analytics workspace where natural-language exploration is the headline interaction. Crucially, the team is pushing AI as an infrastructure layer customers can own — BYO LLM keys, granular access policies — rather than locking customers into a vendor-managed model. The dashboard improvements look incremental, but read as ground prep for AI agents to consume and manipulate dashboards more reliably.
Expect the next quarter to bring agentic dashboard editing — Ask AI not just answering but proposing dashboards and saving them — plus expanded BYO LLM coverage (likely Azure OpenAI or open-weights via OpenRouter) to widen procurement options for enterprise buyers.
See more alternatives to Sigma Computing →
See more alternatives to Holistics →