GovernKit

The mission

Every team should be able to use AI — on their terms. GovernKit exists to make responsible AI adoption faster than the informal process it replaces, not slower. If governance adds friction, people route around it; if it removes friction, it becomes the default.

The inspiration

While building data infrastructure for AI companies, I kept seeing the same pattern inside every organization I talked to: an employee asks in Slack whether they can use some AI tool, someone senior replies “probably fine,” and nothing gets written down. Six months later nobody remembers what was approved, under what conditions, or why.

That informal process works right up until it doesn’t — a customer asks for your AI vendor list, a regulator asks how you vet tools, or someone pastes sensitive data into a chatbot nobody knew existed. GovernKit was born from watching that failure mode repeat everywhere.

What it does

• Structured requests. Employees ask for AI tools in a two-minute form instead of ad-hoc messages, so every request has an owner, a use case, and a status.
• Automated vendor vetting. Privacy policies and terms of service are analyzed automatically, with risk ratings and verbatim citations — no more reading forty pages of legalese per tool.
• AI-drafted policies. Generate a workplace AI policy tailored to your organization and track acknowledgments across employees and contractors.
• Audit-ready by default. Every decision is documented, with dashboards for tool status, risk levels, and AI spend across the organization.

The design principle

The bar for every workflow in GovernKit: it has to be faster than the Slack thread it replaces. A structured request that takes longer than typing “hey, can I use this?” is a failed design. That constraint shapes everything — from the two-minute request form to one-click policy acknowledgments.