We initially looked at AI-FW because we were trying to understand how to put proper security and governance around AI without slowing everything down.

The UI has been easy to work with, and I like that I can quickly see what's happening rather than having to go hunting through logs.

Getting it integrated was also fairly painless. Performance was something I was worried about because the traffic is being inspected, but so far that hasn't been an issue for us.

Where I've found the most value is having security, AI-DLP, governance, monitoring and agent identity together. It saves us from having to think about each problem completely separately, which is also where I see the ROI.

The onboarding experience was great. Securetron included onboarding as part of the service which really help getting things configured and could ask questions as we went along. The training wasn't just "here's where you click". We also spent time understanding the security side of AI and how different policies could be applied, which I appreciated.

The part I find most interesting is the semantic and intent analysis and then their defence in depth approach which drastically reduced false positives compared to other vendors that kept dropping non-malcious traffic. AI interactions aren't always easy to classify using traditional rules, so having more context around what a prompt or response is actually trying to do is useful.

Overall, it's been a good experience. I especially like that AI-FW is tackling some of the newer problems around AI agents, identity and governance, rather than focusing only on basic prompt filtering. Review collected by and hosted on G2.com.

My main downside is that there’s a lot to explore. The basic setup was straightforward, but understanding all the security, governance and AI features took some time. Support was helpful in getting us through that. Review collected by and hosted on G2.com.