Observe.AI

Top three standout features: - QA Automation Logic -- The tool is incredibly robust in the logic that powers auto QA, leveraging keyword spotting, metadata, machine learning, and other internally tracked data points to score questions. The level of specificity and grouping of these different elements allows for nuance in the answer selection process that really allows OAI to be as accurate as possible. - Reporting & Analysis Features -- The tool has a plethora of different data points that can be pulled into reporting, allowing the tool to quickly present many different facets of QA performance on both the agent and analyst level. The Analyze tab in particular is a great standout as it has helped us dig into how often different behaviors/incidents have occurred and create moments to track them. - Calibration -- The tool's way of organizing interactions for QA and leadership to calibrate on the scoring has been a huge quality of life improvement. We're able to very easily ensure everyone is grading on the same evaluation form for the interaction and see monthly trends as we review where people were aligned and misaligned.

There are no outages and their engineers provide reasonable turnaround times for custom work.

Since implementing Observe AI , we've realized significant, measurable improvements across efficiency, coverage, compliance and customer experience. We love the fact that we have been able to increase the number of evaluations by 300% while simultaneously achieving 30x efficiency gains and saving more than 12k hours permonth previously spent on manual QA evaluations. These gains allowed us to scale quality oversight without adding operational burdens such as staffing overhead.

Top three areas of opportunity: - Proactivity and Thoroughness of Tech Support -- While our relationship with the technical support team has improved and we've seen faster turnaround and more adequate responses to issues, there are still some tickets that feel as though the issue (and supporting evidence provided) has been skimmed through rather than thoroughly reviewed, analyzed, and utilized for testing/troubleshooting. I am uncertain if this is a volume issue or if something else is at play, but the technical support team could stand to ask additional probing questions of their users and ensure they understand the full scope of the issue before providing an answer that distills down to things are working as expected. - Lack of Consolidation -- The product as a whole seems built with the philosophy of additions and expansions being separate and loosely tethered together. It would be nice to see more strategic introduction of new features and tools the product offers so that certain elements (like questions on evaluation forms or changes to evaluation forms) could be consolidated rather than having to simply start from scratch and undo/disable/remove with every feature when making changes. - Onboarding and Training Materials -- Throughout the integration and implementation process, we were repeatedly told that high quality onboarding was a point of pride for Observe.ai. It would be nice to see more thought given to the timing of OAI-led training and materials to supplement those trainings as we performed them early on in implementation, before the tool was being fully leveraged. This has resulted in us having to build additional content and provide retraining as the timing was too far apart from when our internal stakeholders were actually using the tool.

My experience with the Customer Relationship Management team, aside from some reps, has been challenging. Interactions often feel overly sales driven rather than collaborative and at times the communication style can come across as dismissive. Product negotiations have been particularly frustrating due to changing terms and perceived bait and switch tactics, which undermines trust. It all seems very male dominated.

At this time, there are no specific aspects of the platform we dislike. Observe AI is highly configurable and can be designed to align closely with our business needs.