Hey everyone! 👋 Just wrapped up the lit review chapter for my CS dissertation (topic: observability in serverless architectures, naturally 😉), and I used ResearchRabbit as my primary discovery tool. Wanted to share my real-world results and some workflow thoughts for anyone considering it.
**The Good:**
* **Visual discovery is a game-changer.** Starting from a couple of seminal papers, the "similar work" maps and author networks helped me find connections I would have totally missed with just keyword searches. It's like building a dependency graph for your research.
* **The feed and alerts are super useful.** Once you seed your collection, the daily feed of new, relevant papers felt like having a CI pipeline for academia. Got notified about three super-relevant pre-prints I'd have found months later.
* **Collaboration was smooth.** Shared collections with my advisor easily. He could comment on papers directly in the app, which streamlined our sync-ups.
**The Gotchas & My Setup:**
* It's fantastic for *exploration*, but you still need your old tools for *precision*. I'd find a great paper in ResearchRabbit, then export its BibTeX and pull the PDF through my university's library portal. The in-app PDF viewer is okay, but I still did my actual reading/annotating in Zotero.
* The "priority" and "exclude" features are crucial. My first map was a mess because it included a lot of adjacent work (like general microservice tracing). Once I excluded a few key terms, the recommendations got much sharper.
* It can get "stuck" in a niche. You have to consciously seed it with papers from slightly different angles or sub-fields to keep the recommendations diverse.
Overall, I'd rate it like a great monitoring dashboard: it gives you the topology and the real-time alerts, but you still need to drill down into the individual traces (papers) with your other tools. It drastically reduced my initial discovery phase and probably added about 10-15 crucial sources I wouldn't have found otherwise.
For anyone starting, my workflow looked like this:
1. Drop 5-10 cornerstone papers into a new ResearchRabbit collection.
2. Explore the map, add promising finds to the collection.
3. Let the daily feed run for a week, adding more.
4. Export all metadata via BibTeX.
5. Import into Zotero for deep reading, notes, and citation.
Hope this helps! Would love to hear how others have integrated it into their workflows.
Dashboards or it didn't happen.