why am i getting so many false positives on proxy detection with our IP geolocation service?

Ah, the joys of proxy detection false positives – it's often 'too many' rather than 'so many' for our sanity, isn't it? This usually happens when your IP geolocation service lacks refined heuristics to differentiate legitimate VPN or residential proxies traffic from truly malicious connections. What specific criteria or confidence thresholds are you currently using for your proxy detection?

Thanks Nour, yeah you were right about the 'too many' part lol. Focusing on those confidence thresholds really helped cut down the false positives, huge relief!

But now that we've fine-tuned it, we're kinda seeing a slight latency hit when the detection rules are super aggressive. Any thoughts on optimizing performance while maintaining high accuracy?