Launched this week
detectly
Turn your fleet's dashcam footage into driver safety scores
70 followers
Turn your fleet's dashcam footage into driver safety scores
70 followers
Fleet managers have dashcams on every vehicle. Almost none of the footage ever gets watched as there simply isn't time. We built detectly to fix that. Upload a clip and the AI automatically flags dangerous incidents: following too close, cut-ins, pedestrian hazards, and more. Every driver gets a risk score. Annotated clips are ready to review in minutes. No manual review. No new hardware. Works with any dashcam footage you already have.
detectly
Hey Product Hunt 👋
I'm Kieran, I built detectly solo, so any feedback lands directly with the person who can do something about it.
The thing that stuck with me when I started talking to fleet managers is that the first anyone hears about a problem driver is usually an accident. The footage exists, but nobody has time to sit and watch hours of it on the off chance something bad happened.
That's the gap detectly sits in. 3 weeks live, 20k social views, and the conversations I've been having with fleet managers since have only reinforced it.
If you manage a fleet or even just a handful of vehicles, I'd love to hear what your current safety process looks like 🙏
detectly
Quick update on detectly for anyone checking this out today.
I’m a solo founder and built the entire thing myself, the frontend, backend, auth, billing, and the ML pipeline that actually detects the near-misses. It runs YOLOv11 for object detection, ByteTrack for tracking vehicles and pedestrians across frames, and a rule-based risk engine that flags incidents like tailgating, cut-ins, hard braking, pedestrian hazards, and lane drift.
The bit I’m most proud of is that it works with footage from any existing dashcam. Most competitors in this space require fleets to install new proprietary hardware and sign long contracts. detectly just needs the footage you already have.
I’ve spent the last few weeks talking to fleet managers directly and the most common thing I hear is that near-misses happen constantly but almost never get reported or reviewed. That’s the exact gap detectly is built to close.
Would genuinely love to hear what people think, what questions you have, or what would make this more useful for your own use case. I’m here all day and will reply to everything 🙏
@detectlykieran that is great insight to hear. It seems like the common approach in the industry when it comes to incidents is reactive rather than proactive. Personally I'm interested to see how detectly helps transform this approach and close the gap between footage and insights.
detectly
@dnghngoc Really appreciate that, you've put it better than I have honestly. That's exactly the shift we're trying to make, footage capturing the moment versus footage actually being used to prevent the next one. Would love to have you try it out and see what you think!