Dreambase is a significant player in the agentic analytics space because of its deep integration with the Model Context Protocol (MCP). By serving as an agentic layer over Supabase, it provides a practical template for how reasoning agents can interact with structured production data. Their "Analyst Agents" are not merely wrappers for LLMs; they utilize a proprietary semantic layer to bridge the gap between natural language intent and complex SQL execution.
For the broader ecosystem, Dreambase represents the shift from passive dashboards to active data agents. Their support for remote MCP connections allows their agents to act as a bridge between a company's primary database and its SaaS stack. This makes them a key component for developers building agentic workflows that require real-time access to production data and external business context without the overhead of a traditional data warehouse.
Dreambase is built on the premise that the distance between a production database and a business dashboard is currently too wide. For most startups, bridging that gap requires a chain of human intervention: a data engineer to move data, a data analyst to write SQL, and a product manager to design the visualization. Founded in early 2025 by Kyle Ledbetter and Andy Keil, Dreambase attempts to collapse this pipeline into a single agentic interface. Based in Texas, the company operates as an official Supabase partner, targeting the growing population of developers who use Supabase as their primary backend.
What makes Dreambase distinct from traditional business intelligence tools is its rejection of the ETL (Extract, Transform, Load) model. Instead of moving data into a separate warehouse like Snowflake or BigQuery—which introduces latency and cost—the platform connects directly to the production Postgres instance. This "database-first" approach is enabled by what they call an AI-native semantic layer. This layer identifies schema relationships and product metadata to give LLMs, such as Claude 3.5 Sonnet, the context needed to write accurate queries. This focus on context is a direct response to the primary failure of early text-to-SQL tools: their tendency to hallucinate column names or misunderstand database normalization.
The core of the product is the Dreambase Analyst Agent. These agents are tuned specifically for the Supabase ecosystem, understanding common table structures and Postgres best practices. Users interact with these agents in natural language to generate reports, which can then be pinned to permanent dashboards or converted into presentation-friendly slide layouts. For teams that need to monitor specific metrics, the platform includes automated "Report Cards" that assess the health and security of a Supabase project, offering recommendations for performance tuning without requiring manual SQL audits.
Perhaps most relevant to the current AI landscape is Dreambase's early support for the Model Context Protocol (MCP). By utilizing MCP, the platform allows its analyst agents to pull in context from external APIs—such as Stripe for revenue data or HubSpot for CRM leads—and query them alongside the core database. This turns the analytics dashboard into a reasoning hub where an agent can correlate production data with third-party business signals. The pricing model, which ranges from $29 to $249 per month, positions the tool for the "prosumer" and mid-market startup segment—teams that are large enough to have complex data needs but too small to justify a dedicated data science department. By focusing on a specific technical ecosystem rather than attempting to be a general-purpose tool for every database, Dreambase provides a plug-and-play experience that incumbents often struggle to match in terms of setup speed.
AI-native analytics and reasoning agents for Supabase and Postgres.
Dreambase is hiring