In July 2026, PromptQL released its AI-native workspace, positioning itself as a "multiplayer AI" alternative to Slack. The product, built with Hasura, React and Tailwind CSS, lets teams create shared AI threads where prompts, answers and contextual data are captured in a living knowledge base.
The launch was announced on X by co‑founder Tanmai Govindan (@TanmaiG), who described PromptQL as "the world's first AI native workspace" and credited the accidental development of the product to the team behind Hasura. He wrote:
"The team behind PromptQL is Hasura, and we ended up building PromptQL entirely by accident."
Co‑founder Rajoshi Ghosh added that the recent rollout of GPT‑5.6 (referred to as "Sol") shifted PromptQL's ambition from delivering better answers to acting as a substrate for multi‑step, production‑grade tasks. In a follow‑up X post, Ghosh explained that the new model enables teams to hand off complex feature development to AI agents without constant supervision, turning shared context into executable workflow.
Kevin Systrom (@kevin) — listed in the company’s launch post — celebrated the release, noting the product’s focus on "captured team context rather than just isolated chats" and encouraging early adopters to try the platform.
PromptQL’s core proposition is a shared context layer that records every interaction, decision and data point across the team’s tooling ecosystem. Users can tag teammates within a thread, prompting the AI to pull in data from connected databases, SaaS services (e.g., Customer.io), and internal APIs. The platform claims to automatically generate connectors on demand, allowing AI agents to execute tasks such as building integrations, writing code, or generating reports.
During the launch, the team ran a public beta with a 70‑person internal crew that migrated from Slack to PromptQL in February 2026. According to Ghosh’s X comment, the experiment triggered what he described as "team‑wide AI psychosis" as the shared AI began aggregating information from across channels, repositories and project boards. The team interpreted this as proof that the product had effectively become a "Slack for the AI era," eliminating the need for direct messages or meetings because the AI could surface real‑time status updates.
Early users reported a noticeable reduction in Slack traffic. One beta participant, identified as @nk_kapur, praised PromptQL’s ability to store cross‑functional context—research, marketing, finance and product decisions—in a single place. The user highlighted the platform’s rapid access to the latest LLM models, noting that they could test new models "almost the day they are launched" and compare performance on internal workflows.
PromptQL also supports external collaboration. Following user feedback, the team rolled out "external teams connect v1," enabling guests such as consultants, vendors, or lawyers to join private threads while preserving scoped permissions. Permissions are managed at the integration level: personal OAuth connections default to the individual user, while database connections can be shared with team‑wide read or write restrictions.
Security and privacy surfaced as recurring concerns in the X discussion. Users asked how the platform prevents accidental exposure of sensitive APIs when AI agents forge new connectors. Co‑founder Ghosh replied that each integration respects the owner’s access controls, and that developers must explicitly grant permissions for any data source. The company’s public documentation—linked from the product’s homepage promptql.io—details role‑based access and audit logs, though the depth of these safeguards remains to be independently verified.
PromptQL is offered as a SaaS subscription with a free tier for early adopters. The launch announcement included a $1,000 token credit for Product Hunt users willing to trial the platform with their teams. The company also promises a roadmap that includes recurring workflow templates, admin usage dashboards, editable tables within the AI canvas, and tighter integration with monitoring tools like PostHog and Grafana.
While the product’s ambition is clear, its market positioning raises questions. Slack, Microsoft Teams and emerging AI‑augmented collaboration tools such as Microsoft Copilot for Teams already embed LLM capabilities. PromptQL differentiates itself by treating the entire workspace as an AI agent that can generate task‑specific agents on the fly, rather than merely supporting human‑to‑AI chat. Whether this architectural shift can overcome entrenched user habits and the network effects of established platforms remains to be seen. The launch marks PromptQL’s first public offering after a year of internal testing. No funding round, valuation or investor involvement was disclosed in the announcement, suggesting the company is bootstrapped or early‑stage. The team’s background in building Hasura—a popular GraphQL engine—provides credibility in handling real‑time data pipelines, a critical component for the promised on‑the‑fly connector generation.
PromptQL’s emergence reflects a broader industry trend: tooling that merges collaborative communication with AI‑driven workflow automation. By embedding LLMs directly into the knowledge base, the startup aims to transform how teams capture, retrieve and act on institutional memory. Whether the platform can scale beyond its initial beta and attract enterprises that currently rely on Slack’s extensive ecosystem will determine its long‑term impact.
Key takeaways
- PromptQL launched as an AI‑native workspace that replaces Slack‑style chat with shared, editable AI threads.
- Co‑founders Tanmai Govindan (@TanmaiG), Rajoshi Ghosh and Kevin Systrom (@kevin) positioned the product as a "multiplayer AI" platform built on Hasura, React and Tailwind.
- Early beta users report reduced Slack traffic and faster cross‑team knowledge retrieval.
- Security model relies on per‑integration permissions; details are limited to product docs.
- No external funding disclosed; the venture appears self‑funded or pre‑seed.