технология анализа голоса Nemesysco eng
•Download as PPTX, PDF•
0 likes•419 views
Nemesysco develops voice analysis technology that can identify hidden emotions and assess truthfulness by analyzing uncontrolled properties of the voice. Their Layered Voice Analysis technology can detect various emotional states. They apply this technology to products like LVA-i for pre-employment screening and integrity assessments of employees. LVA-i balances accuracy, time, and cost compared to alternatives like polygraphs. It uses questionnaires and automated reporting to assess truthfulness and suitability of candidates. Nemesysco cites case studies where their technologies helped reduce fraud, customer complaints, and call times at various companies.
More Related Content
Similar to технология анализа голоса Nemesysco eng (20)
More from MMI Group (19)
Recently uploaded (20)
технология анализа голоса Nemesysco eng
- 2. What is Nemesysco’s Voice Analysis Technologies all about? It is NOT about what we say or how we try to sound… It is about how we truly feel about it
- 3. Nemesysco Voice Analysis Technologies For the past 17 years We’ve been researching and developing a unique emotion analysis technology based on proprietary measurements of the human voice.
- 4. Nemesysco Voice Analysis Technologies Our voice data was collected from various real-life sources and controlled experiments from all over the world in order to insure our technology is language independent.
- 5. People can’t control their genuine emotions as they are expressed in the voice. Nemesysco’s technology is unique in it’s ability to identify the unexpressed and genuine emotions as it is based on uncontrolled properties of the voice. Uncontrolled expressions Vs. Acted ones
- 6. Nemesysco Layered Voice Analysis (LVA) Nemesysco provides various solutions powered by “Layered Voice Analysis” (LVA) technology SENSATIONS: Positive, Conflict, Negative Stress, Excitement, Hesitation, Confusion, Mental Effort Concentration, Embarrassment, Energy, Profound Emotions Arousal, Depression Anger, Sadness, Happiness, Clinical States
- 7. Fraud Detection Security HR Screening Call Center Nemesysco’s products powered by LVA technology
- 8. Integrity Assessment Systems For HR Screening
- 9. On a daily basis HR managers need to choose between: Why LVA-i? highly reliable, but uncomfortable and expensive evaluation procedures low cost testing, but to expect far less reliable results. to Polygraph Pen & paper LVA-i Accuracy Time Price Integrity Assessment Systems For HR Screening
- 10. Why LVA-i? On a daily basis HR managers need to choose between: highly reliable, but uncomfortable and expensive evaluation procedures low cost testing, but to expect far less reliable results. to Polygraph Pen & paper LVA-i Accuracy Time Price Integrity Assessment Systems For HR Screening
- 11. . Immediate availability . Optimal accuracy . And a fair price tag Polygraph Pen & paper LVA-i Accuracy Time Price Nemesysco’s LVA-i balances between:
- 12. LVA-i is a vetting and assessment platform that integrates our unique voice analysis technology with carefully designed sets of questionnaires. What is LVA-i?
- 13. LVA-i has two applications 1 2 Pre-employment The initial phase of the recruitment procedures. Periodical Periodical test for all levels of employees – organizational “honesty maintenance”.
- 14. The LVA-i’s assessment process 1 2 3
- 15. Emotional reactions during an integrity assessment Nemesysco’s technology will be most effective in detecting emotional reactions under the following 3 conditions: When the interviewee is motivated to convince When the interviewee knows that he is not telling the truth and he has an intention to deceive When the interviewee has something to lose if caught lying 1 2 3
- 16. LVA- i Automated Report Logic
- 17. Personal and current History 3rd Party Background attitude LVA-i questionnaire Funnel approach
- 18. Confidentiality and Secrecy Theft from your place of employment Bribery and Kickbacks Fraud and Deceit Team work Drug usage Alcohol Consumption Gambling Loyalty to the company LVA- i Standard questionnaire’s topics
- 19. Risk test Personality test Management skills test Job specific test (Drivers, Cleaners, guards) LVA- i tailor made questionnaires
- 20. LVA-i has 2 modes of operation 1 2 Over the phone Structured interview assisted by a phone operator Interactive post Automated interviewing procedure in a Man-PC interaction
- 21. LVA-i is friendly and intuitive… No age, gender or ethnic bias Language independent Automated report generation Intuitive and easy to understand reports Enables mass vetting in short time frames
- 22. LVA-i Reports
- 23. Case study Castel Detect™ - NORTHSTAR Companies Northstar relies upon Castel Detect™ for monitoring live conversations, detecting incidents of business risk and measuring agent performance. 54% Decrease in Escalated Customer Calls Today, Castel Detect™ provides Northstar agents the ability to easily identify emotionally charged calls. Agents are successfully reengaging customers to deliver successful resolution and customer service during live calls. 63% Decrease in Customer Complaint Letters As not all customers request to speak to a supervisor at time of issue, customer complaint letters, too, have dropped at Northstar. Not only is this a positive result demonstrating first call resolution, but an additional benefit in reducing time invested managing customer letters. 35% Decrease in Call Handle Time The technical correctness of a call for Northstar includes compliance, successful solution, business commitment and customer satisfaction. Each is vital to Northstar’s corporate philosophy. Ensuring customers are provided necessary disclosures and given suitable remedies all while receiving best in industry customer service requires appreciation and strategic use of one’s time. Since employing Castel Detect™, agents have shown increased communication effectiveness resulting in decreased call handle time.
- 24. Case study MTS - Mobile TeleSystems OJSC (Russia, 2006) Method • Ran analysis on 81,000 calls • Thoroughly researched - 3,000 calls • Compared QA5 analysis to MTS contact center supervisors analysis (same calls) Findings • 77% compliance between QA5 analysis and MTS supervisors on all calls (marked “Good”/ “Problematic” / “Bad”) • 95% compliance on marking of “Good / Normal” calls Conclusion For off-line quality check • Listening to 3% of the QA5 marked calls is enough for finding 30% of bad calls handling by operator • Listening to 1.8% of the QA5 marked calls is enough to find 40% of bad calls by customers • Using QA5 can reduce by up to 4 the time supervisor spends listening to calls and allows him to listen to up to 9 times more calls *MTS is the biggest cellular operator in Russia, with over 90 million customers
- 25. Case study RA7 Fraud prevention solution at UNIQA Insurance (Hungary, 2011) Sample incident: a 4.3 magnitude earthquake in Hungary, Europe. Saturday, January 29, 2011 at 17:41 The day after the earthquake: 163 claims were submitted over the phone. • The further away from the epicenter of the quake - the more fraudulent claims were recorded • Operators in the call center were able to handle and defeat fraudulent claims as they were received by phone • Only 119 claims were found to be truthful & 44 (27%) to be fraudulent Magnitude Legitimate claims High Risk claims 4.3 99 2 3.6 20 18 2.2 0 18 1.3 0 6
- 26. Case study Domodedovo International Airport (Moscow, 2005) Method The focus of the pilot was to ascertain the feasibility and accuracy of the GK1 in the detection of drug smugglers entering Russia via the airport. The local RDEA authorities selected a number of passengers from each flight. These selected passengers were tested by the GK1 Standard Post system - they were then tested by the GK1 Online Post and then finally, were physically checked for possession of drugs. Findings • 11.6% of the passengers tested were found by the automatic questionnaire to be Red. All of these passengers were then cleared by real-time interview. These numbers were confirmed by physically checking the passengers • There is 100% accurate match between the results given by the GK1 system and the actual physical examinations Gree n Inconclusiv e Red Automatic questionnaire 437 1 58 Real-time interview 495 1 0 Physical check 496 --- 0 • In one case, the GK1 system indicated that a passenger is somehow connected to Drugs. We were later informed that this passenger was in fact a drug dealer, and that police intelligence had already marked him • No Red passengers were found during the pilot period
- 27. MMI Group 8 Illinska st., 3rd Entrance, BC "Illinsky“ Kyiv, 04070, Ukraine tel.: +38 (044) 237-70-81 e-mail: info@mmigroup-it.com www.mmigroup-it.com Thank you! IT-innovations. Strongly recommended!