Dreambase raised $3.7 million in funding for its AI-powered analytics platform that eliminates the need for dedicated data teams.

The Austin-based startup has developed AI data agents that perform analytical work traditionally handled by human data teams. CEO Andy Keil and CTO Kyle Ledbetter launched the company in August 2024 after months of prototyping, offering early access in April 2025.

Dreambase describes itself as Supabase-native rather than merely compatible. Users can connect their Supabase database and generate dashboards "in seconds," according to Keil.

"Our AI data agents handle the rest," Keil said. "They build dashboards, run analysis and surface insights 24/7. Whether you're a solo founder shipping your first product or an enterprise AI innovation team running mission-critical analytics, you get a full virtual data team without the headcount."

Supabase executives join the round

Three Supabase executives — the CFO, CTO, and COO — participated in the funding round as individual investors. Felicis led the raise, with participation from Active Capital, FirstMile Ventures, Darkmode Ventures, Angel Collective, Earl Grey Capital, and Mercury Fund.

Angel investors from Perplexity, Cloudflare, and QuotaPath also backed the round.

While Dreambase's technology works with any Postgres database, the company sees significant opportunity within Supabase's 7 million developer community. Keil said the startup plans "to be the analytics layer for every company running on Postgres in the world."

The platform provides a "context layer" to large language models, allowing users to ask product questions in natural language directly against their database.

Felicis General Partner Viviana Faga said her firm was impressed that Dreambase's founders had "lived the problem" from both sides — Keil as head of product at QuotaPath and Ledbetter in design roles at Teradata, MicroStrategy, and eBay.

Supabase CFO Scott Buxton said he invested because Dreambase "will win the analytics layer for the next generation of Postgres-native companies." He noted that most analytics tools treat Postgres as "just another data source to extract from," while "Dreambase treats it as home."

The five-person team plans to double headcount over the next few months, hiring across engineering and enterprise go-to-market functions.