RankMiner Predictive Voice Analytics Platform enables Call Centers to do to three things… identify which customers are at-risk, which customers will buy more, and which agents may help, or hurt, their business.
“Unlike Speech analytics… RankMiner’s approach is a game-changer”
“I loved how RankMiner was able to predict which customers were most likely to pay.”
“RankMiner’s Agent Insight models have increased our QA productivity by over 50%”
“I didn’t realize how much money was left behind… the result opened my eyes. It definitely works!”
Our patented and proven Artificial Intelligence algorithms help Call Centers to drive more predictable business outcomes directly from voice calls.
The RankMiner Approach: Direct-to-Ops Predictive Modeling
RankMiner takes a fundamentally different approach to traditional speech analytics solutions; our primary concern is in predicting future outcomes. We do not simply generate more data to provide company analysts with a “jumping off” point for deeper study. We start with the end-game in mind. What does the company want to see more of? Increased sales? Improved customer experience? Greater spend from existing clients? What does it want to see less of? Reduced agent turnover? Reduced customer attrition?
RankMiner’s approach is different – unlike Speech Analytics, we correlate voice-based emotions and behaviors to specific business outcomes, creating “vocal imprints” that are used to predict future outcomes, therefore eliminating the requirement of human observation & hypothesis testing, a necessary process to utilize traditional speech analytics solutions.
RankMiner’s unique feature extraction technology adapts techniques similar to those used in facial recognition. Facial recognition technology exploits symmetry and localized characteristics to identify various facets of the image. Speech signals can be analyzed in much the same way, because its frequency spectrum can also contain symmetry and contextually adjacent indicators.
Most audio analysis applications rely solely on pitch-based features in a fixed frame of reference. Unfortunately, these coarse measurements are not always effective at identifying the same emotions across the various vocal ranges. RankMiner’s proprietary algorithms look beyond the fixed frequency scale and calculate features that describe speech in a relativistic sense, in essence capturing the variations within an individual’s vocal patterns and isolating the traits shown to differentiate emotions across a variety of people. RankMiner’s custom-designed predictive models then use all of these innovative features to differentiate speech from music or noise and identify emotional content within the speech. Being able to infer emotional patterns about entire populations significantly boosts predictive power and enables RankMiner products to work across different languages, cultures, and other communication barriers. The increased predictive power translates directly into increased accuracy in predicting behavior-based future outcomes. This is in stark contrast to using the speech signal to increase accuracy in translating to text, then expecting the translation to “speak for itself.”
Fuses the link between emotional response and customer/agent activity without the need to first translate and analyze the dialogue.
Can process data, extract features, generate predictive analytics, and deliver prescriptive reports for a high-volume call center using an off-the-shelf PC.
Innovative methods for feature extraction and predictive modeling across languages.
Extremely capital efficient development team with expertise in mathematics, predictive modeling, machine learning and signal processing.
Preston Faykus – CEO
Preston is currently CEO & Founder of RankMiner Predictive Analytics, a company that analyzes voice-based emotions and behaviors to predict human behavior. Call Centers use RankMiner’s patented machine-learning technology to systematically increase profits by predicting future business outcomes and improving agent and customer success.
Starting in Moscow, he spent the first 10 years of his career building businesses across Central and Eastern Europe. Prior to RankMiner, Preston held senior management positions with DialogBank, EuroNet Worldwide, Aurum Technology and FIS where he focused on growing the companies’ electronic payments and data analytics business.
Preston holds an MBA from the University of Chicago and an undergraduate degree in Finance from the University of Texas at Austin. He lives in Saint Petersburg, Florida with his wife and three kids, where he has held multiple leadership and board positions including First United Methodist Church, the Rotary club of Downtown St. Pete, Studio 620 and other organizations. He loves to cook, play tennis and spend time with his family.