Frequently asked questions

Yes, you can, and I'd even be happy to help you doing that. However, you risk that without really good prompts, skills and other context, you will end up with a very colorful report, that you will spend a lot of time on, and later will have difficulties using with customers, auditors or regulators, as it is not built to industry standards, or compliant with your company's risk frameworks.

Huginn is not better, it is just different. Huginn was built by risk management experts, and offers a very strong human oversight, which is important. Risk reports can get sensitive, and you don't want an LLM getting too creative. Huginn is "tamed" in the sense that it only works within the risk taxonomy it receives and follows the exact same process a human would.

Huginn started as a research project about 3 years ago, I needed to learn AI risk management, and decided to learn by doing. So I (vibe) coded it. The primary thesis I wanted to test- can LLM's do proper risk management and how can we use AI for doing our jobs? Ultimately, can AI manage risks with other AI's, was also intriguing me. The learnings have been implemented into Huginn, and I'm now sharing them. It is a thrill of a journey, and it only just started. I will be writing about the Huginn journey on LinkedIn. Please connect.