There is a quiet, exhausting reality at the heart of every modern enterprise. Despite spending millions on massive data warehouses and modern infrastructure, actually getting a simple answer about your own business remains agonizingly slow. Business teams are forced to wait in weeks-long ticketing queues for data engineering teams to write SQL. And when companies try to solve this by pointing generic AI at their data, they immediately hit the “confidently wrong” problem — hallucinations caused by missing tribal knowledge and broken security guardrails. We have built an incredible amount of data infrastructure, but the access layer is fundamentally broken.
It takes builders with profound infrastructure rigor and the courage to completely reinvent their own success to solve this at scale. That is exactly what Tanmai Gopal and Rajoshi Ghosh are doing with PromptQL.
Founded originally in 2017 as Hasura — the legendary open-source GraphQL engine downloaded over 400 million times — Tanmai and Rajoshi have made a massive, visionary pivot. PromptQL was founded in 2024, and they are building the definitive AI data agent that creates and executes complex query plans on business-critical data.
PromptQL connects directly to an enterprise’s existing data infrastructure, whether that is Snowflake, Postgres, or MongoDB. But instead of just chatting, it deploys what Tanmai calls “Multiplayer AI”. It operates as a shared, collaborative workspace where human teams and AI agents work together in a single thread. You don’t just prompt it; you tag it like a colleague, and it writes the code, pulls the data, and delivers the final analysis.
The true moat here is uncompromising accuracy and deterministic security. PromptQL natively enforces permissions at the data layer, ensuring the AI only sees exactly what the user is authorized to see. Furthermore, by capturing the deep, messy tribal context of a company just-in-time as the team works together, its accuracy compounds over time. It completely abstracts away the soul-crushing bottleneck of the data request queue, shifting the paradigm from waiting for dashboards to instantly conversing with your data.
The market? Fortune 500 enterprises, CXOs, and fast-scaling business teams who desperately need reliable, secure AI to drive their operations without betting their company’s privacy on an unpredictable model.
Seeing founders of this caliber head-down, choosing to walk away from the comfort of a billion-dollar legacy to build the spiritual successor to GraphQL for the age of AI, is profoundly inspiring. They are quietly rewriting the playbook for enterprise data access, giving organizations their momentum back.
Let’s celebrate the builders.
w/ Jay Ingle
