Mito AI has been live for less than a month, and since then we’ve watched our open source users use it in incredibly creative ways.

Our users have used Mito AI to work around limitations in our product, do complex transformations with a single request, optimize their scripts, and generally augment the automation capabilities of Mito to a degree we didn’t expect.

To help our enterprise customers accelerate their adoption of cutting-edge automation tools, we just launched a bring-your-own model feature for Mito Enterprise users.

Enterprise Size, Enterprise Problems

Mito AI v1 required an API call to OpenAI’s servers. Moreover, this API call included some of the users data (e.g. column headers) that the code generated by OpenAI was relevant to the users specific problems.

But this made Mito AI v1 a bad fit for our enterprise customers - most of whom have extremely stringent data-privacy policies. As a result, Mito is deployed on premise, and we ensure that no customer data should leave customer systems, ever.

Bring your own model

As of today, Mito Enterprise customers can now bring their own large language models to Mito. With a simple configuration change, Mito AI can now be used with a model that is deployed on premise, ensuring that no data ever leaves our customers systems.

If you’re an enterprise with a large language model you’re looking to deploy to your employees, reach out if you’re interested in safely accelerating adoption. Our spreadsheet and AI integration means that users can effectively evaluate LLM generated code for accuracy more effectively than any other tool that exists currently.

If you’re an enterprise that has not deployed a large language model on premise, but you’re interested in getting the cutting-edge in automation tools — we’ve already begun work to help users adopt these tools - reach out.