Hey everyone! We're a medium-sized resources business dealing with the usual IT headache of ensuring user safety when it comes to handling sensitive documents, especially with the rise of public LLMs. We've got about 200 staff members, but around 50 of them will be heavy users who rely on these tools to rewrite and analyze documents. It's crucial for us to make sure that no sensitive data leaks to these public platforms.
We tried using Copilot, but it didn't meet our needs. Now, we're considering setting up a custom front end for our users, effectively blocking access to public LLMs. We're thinking of using something like Open WebUI and host it on our ESX cluster, managing API access through OpenAI with a fixed budget of tokens for each user.
Also, I'm curious if Azure SSO would be a viable option for us. I've heard about Azure OpenAI, but I'm uncertain about the best direction to take. Can anyone share insights on effective solutions and estimate costs? Just a note that we're based in Australia!
1 Answer
You might want to stick with Copilot if you can implement Enterprise Data Protection. Developing a policy that makes it the go-to choice for everyone could help ease data concerns. It's less about the tech and more about setting the right policies to handle breaches as they come without stifling productivity.
Related Questions
xAI Grok Token Calculator
DeepSeek Token Calculator
Google Gemini Token Calculator
Meta LLaMA Token Calculator
OpenAI Token Calculator