Observe.AI: Technology Must Be Used Utmostly To Strengthen Human Connections

Observe.AI is a voice AI platform that provides the agent with real-time feedback on customer sentiment and guides them on the next best action during the customer call. The AI platform uses deep learning and natural language processing (NLP) to understand the context and generates suggestions and guidance for the agent.

Sharath Keshava Narayana, Chief Revenue Officer, Observe.ai explains the importance of technology in this sector.

How do you think is the contact center market faring amid the pandemic? 

Observe.AI’s Voice AI Platform—which uses speech-to-text and machine learning technologies to help organizations analyze 100% of calls, evaluate, and coach agents—is uniquely helpful for teams that have suddenly transitioned to remote work. Without Observe.AI, most teams manually review just 1-2% of calls and interactions using multiple tools and spreadsheets. Until now, evaluations and feedback were generally given in-person and agents were infrequently coached, but could overhear each other on the call center floor.

Now, agents must receive more feedback and coaching to remain on-message with everyone working from home. 

The key differentiator between Observe.AI and legacy contact center software companies lies in our fast, accurate, and human approach. Our agility as a disruptor places us in a unique position to reach businesses who, like us, must operate at a quick speed in order to grow. Our implementation and activation takes around three weeks, compared to our competitors who have three-month long or greater implementation timelines. We put an entire team of onboarding managers, technical consultants, and Customer Success Managers behind our clients and do not charge any additional services fees, taking a genuinely human approach to customer support.

We’re the only speech analytics solution that has a guaranteed 80% speech-to-text transcription accuracy Service Level Agreement (SLA), meaning we promise to meet at least this level of fidelity in our analysis of calls. Data accuracy is critical to scaling AI use cases, and in areas like sentiment detection specifically, we have seen our platform outperform the accuracies driven by Google and Amazon.

With our tonality-based sentiment detection, many contact centers are able to unlock customer sentiment insights and understand emotions on calls for the first time to better coach agents. Our Voice AI platform draws insights on sentiment, emotion, intent, silence, and more to improve the Customer Experience (CX).

Could you highlight a little on Observe.ai’s Voice AI Platform?

Observe.AI’s Voice AI Platform—which uses speech-to-text and machine learning technologies to help organizations analyze 100% of calls, evaluate, and coach agents—is uniquely helpful for teams that have suddenly transitioned to remote work. Without Observe.AI, most teams manually review just 1-2% of calls and interactions using multiple tools and spreadsheets. Until now, evaluations and feedback were generally given in-person and agents were infrequently coached, but could overhear each other on the call center floor.

Now, agents must receive more feedback and coaching to remain on-message with everyone working from home. 

The key differentiator between Observe.AI and legacy contact center software companies lies in our fast, accurate, and human approach. Our agility as a disruptor places us in a unique position to reach businesses who, like us, must operate at a quick speed in order to grow. Our implementation and activation takes around three weeks, compared to our competitors who have three-month long or greater implementation timelines. We put an entire team of onboarding managers, technical consultants, and Customer Success Managers behind our clients and do not charge any additional services fees, taking a genuinely human approach to customer support.

We’re the only speech analytics solution that has a guaranteed 80% speech-to-text transcription accuracy Service Level Agreement (SLA), meaning we promise to meet at least this level of fidelity in our analysis of calls. Data accuracy is critical to scaling AI use cases, and in areas like sentiment detection specifically, we have seen our platform outperform the accuracies driven by Google and Amazon.

With our tonality-based sentiment detection, many contact centers are able to unlock customer sentiment insights and understand emotions on calls for the first time to better coach agents. Our Voice AI platform draws insights on sentiment, emotion, intent, silence, and more to improve the Customer Experience (CX).

How are you leveraging speech analytics to enhance categorisation, clustering, sentiment analysis, and concept extraction?

Our primary focus is speech analytics: we use AI to access sentiment analysis on voice calls. Observe.AI’s Voice AI engine, an internally developed artificial intelligence (AI) and machine learning solution, leverages speech analytics to enhance categorization, clustering, sentiment analysis, and concept extraction. For example, the system uses machine learning to auto-suggest lookalike keywords and phrases to monitor.

Do you think technology is capable of strengthening human to human interactions? If yes, how?

We believe that as much as possible technology should be used as a way to strengthen human-to-human connections and allow people to reach their full potential, not replace them. Brands are made up of the people who build and represent them. Technologies like AI can and should be used for some forms of automation that free up time spent on manual, tedious tasks (such as searching for and listening to the full length of a call to evaluate it). We believe that through these capabilities, AI can spur business leaders to engage in strategic analysis that tackles deeper employee and customer challenges on a larger scale.

Kindly elucidate on the nature and amount of funding raised by Observe.AI.

We raised $26M in Series A funding in 2019.

What are your immediate and long-term milestones like?

Observe.AI releases in Spring, Summer, and Fall. Planned enhancements include:  

• Agent Coaching 

• • • Observe.AI will release its agent coaching product later this year. Supervisors will be able to use machine learning to surface the most important calls, interactions, and evaluation forms they need to prepare for effective coaching sessions. Machine learning will recommend which agents to coach on what topics and surface up referenceable interactions within calls to play or pull into a playlist of referenceable audio clips for further training. Machine learning will recommend topics that training teams should address with content or programs, as well as track the impact supervisors, training content, and programs have on driving behavior change across various KPIs, such as reductions in Average Handle Time or Negative Sentiment.
  • Real-Time Speech Analytics
    • Observe.AI currently has a real-time speech analytics application in-development that will process and mine the content in live audio streams. User-defined rules will trigger automated actions based on real-time speech events. Automated actions may include sending customized alerts to agents and supervisors, issuing a short message (SMS) or email, launching a document to assist agents, providing context-sensitive information, or automating after-work. 
  • Automation of Agent Evaluations 
    • Observe.AI is currently developing automated scoring of voice interactions utilizing rule-based evaluation templates. Form questions may be related to groups, groups of words, phrases, or sequences of events. As interactions are ingested into the system, they will be run against the corresponding evaluation form to validate the presence or absence of the defined words or phrases and are scored based on the point values associated with each question. Form rules can also be defined to identify which calls should be placed in a supervisor queue for further analysis or agent coaching & training. 

• Future Omni-Channel Analytics Offerings: 

• Observe.AI is developing text analytics applications to analyze content in text-based channels, such as emails, chatbots, and social media interactions. Observe.AI will generate reporting, which may include word clouds, charts, and graphs related to sentiment analysis, silence, and positive/negative/neutral conversations. Analysts will be able to perform configurable proximity searches for words before and after the selected subject, and/or drill down into specific interactions in the data set.

Any piece of advice for the wannabes?

It’s not about ideas. It’s about making them happen—and showing up every day to incrementally grow and progress. We’re looking for people with a can-do attitude who are committed to working toward a vision. Interested professionals, including engineers and product managers, can learn more about the opportunities we have available at observe.ai/careers.