Hi everyone. I wanted to share a recent experience from my university lab that might be helpful for others evaluating Iris.ai, especially in an academic context.
We were genuinely excited to try it. Our research group does a lot of systematic literature reviews, and the promise of an AI tool to help with filtering and mapping was very appealing. We used the full trial period, with about five of us putting it through its paces on a couple of ongoing projects.
In the end, we decided not to proceed with a paid subscription. The main reason was a significant mismatch between the tool's capabilities and our actual workflow. While Iris.ai is good at finding papers based on a broad concept, we found its precision for highly specialized, niche topics in our field (computational linguistics) wasn't strong enough. We spent more time correcting its keyword mappings and filtering out irrelevant papers than we felt we saved. For us, the cost-benefit just didn't add up.
It’s a solid tool, and I can see it being great for interdisciplinary or broader topics. But for our specific use case, it felt like using a sledgehammer to crack a nut—powerful, but not precise enough to justify the ongoing expense. We’ve gone back to a more manual approach combined with very targeted keyword alerts in traditional databases.
Has anyone else had a similar experience, or found a way to configure it for deeper specialization? I’m curious if we might have missed something.
— Eric
Keep it civil, keep it real.