We provide instructions on how to collect relevant data points from each supported server in your infrastructure. Collector agents upload findings into on-prem Collector node that parses and optionally anonymises data. Once ready, parsed inputs are shared with ChatGPT model over API to collect initial findings and eventually generate high (infrastructure overview) and low (server-level) reports with identified issues and proposed remediation steps. Remediations are provided in a form of step-by-step instructions that are generic at first but will get more specific based on your infrastructure. All the specific instructions are stored on the on-prem Knowledge Base.
At this time only parsed and anonymised inputs have to be shared with ChatGPT model via API. Once generated, high/low level reports and remediation instructions are downloaded to be kept on your on-prem servers. Future offerings may require additional uploads to further ehance available services, but any and all such requirements will be communicated well in advance. The plan is to have these features as optional, meaning you will not be prevented from upgrading to the latest version of TechStack.AI suite even if there’s a strict requriement of no uploads of any kind in place.
We treat security very seriously. TechStack.AI suite is a transparent solution that will support automated vulnerability scans in the near future. This means you’ll be receiving full list of ChatGPT prompts that can be scanned and verified prior to activating them.
All the API calls will be made using your customer-specific OpenAI API credentials - meaning no other customer or TechStack.AI agent could access or shared them. Although management of OpenAI API keys is entirely up to the customer, our strong recommendation is you obtain a set of OpenAI API keys specific to TechStack.AI suite.
All the API calls will be made using your customer-specific OpenAI API credentials - meaning no other customer or TechStack.AI agent could access or shared them. Although management of OpenAI API keys is entirely up to the customer, our strong recommendation is you obtain a set of OpenAI API keys specific to TechStack.AI suite.
There are no automated fixes provided right now. Expectation is that TechStack.AI reports will provide valuable guidance to your internal helpdesk teams but actual remediation is outside of the core offering right now. It is absolutely possible to discuss and come up with automated approach once you receive high enough number of identified issues and confirm provided guidance (step-by-step instructions) as acceptable.
Yes, we welcome feedback and feature requests from our customers.