Before and after analysis shows that mean time to detect anomalies in vehicles came down significantly. It is not only mean time to detect but mean time to respond as well to incidents and containing the incident that came down as well. Fleet visibility coverage went on to become significantly higher percentage connected vehicles being monitored. Cost avoidance occurred because of proactive detection, defects were caught early, and warranty claim reduction came down. There are operational ROIs, warranty ROIs, compliance ROIs, deployment and operational ROIs, and business risk ROIs. We could measure cost avoidance, efficiency gains, revenue protection, and various things on which we could measure the ROI. Areas of improvement or enhancement that could be considered are probably a few. Upstream Security can provide deeper native visibility in ECU level behavior. Reduction in false positives is possible because even though the AI and ML detections are very powerful, sometimes security teams want more explainability, and alert fatigue can occur, which is the case with most platforms, nothing very specific to Upstream Security. There could be strong root cause correlation across IT, vehicle, cloud, more improved attack path mapping across telemetrics, and more predictive security. If there is one thing that we would possibly want Upstream Security to look at further enhancing, it is moving beyond detection towards analysis and analytics going forward. This is what most of the OEMs would be looking for. Upstream Security is a great platform and nothing is apparently or fundamentally missing. We expect more visibility, more reduction of false positives, better explainability, simpler dashboards, stronger root cause analysis, strong ecosystem integration, and so forth. This is the case with every platform, and there is no platform that would rate as 10 on 10 because there is always scope for improvement. Our rating of nine marks all the boxes, but the continuous improvement aspect has taken away one out of that number 10.
