If there is one aspect students have aced over the years, it is to keep a straight face while the teacher is explaining yet be in their dream world. This was further enabled with virtual schooling that allows students to protect their privacy and opt not to turn on their cameras. What if one were to tell them their teachers can now understand what a student is doing and read their emotions? Welcome to the growing world of sentiment analysis through AI.
By definition, sentiment analysis is an analytical technique that uses statistics, natural language processing, and machine learning to determine the emotional meaning of communications. The technology is used for several cases, including evaluating response to stimuli, such as in advertisement calls, schools, social media posts and more. In addition, with the increasing move to virtual meetings and calls, companies are finding it more accessible to leverage such tools.
Sales analysis
Uniphore is one such tool using AI to help sales personnel understand the response from customers or potential consumers better. The platform combines computer vision, voice AI and tonal emotion to understand conversations. Uniphore’s proprietary system, Q for Sales, studies ‘emotional intelligence for revenue teams’. The solution offers EQ-based sales real-time and post-call sentiment, engagement, and EQ-based conversation intelligence for revenue teams. This would monitor signals such as the customer’s engagement level increase on hearing about a particular product but flatten on knowing its price. Companies can then use this data to price or situate themselves accordingly. It labels the potent as ’at risk’, ‘on track’ and ‘exceeding’. As per a demo video posted by the company, it will also tag the user’s emotional score, satisfaction, activation, engagement level, and a scale of moods for happiness, surprise, sadness, anger, and more.
Source: Uniphore
Sybill, a similar platform, claims to “Read your prospects’ body language on Zoom calls to help you understand them better. Turn your prospects into customers with an X-ray vision of their needs.” Like Uniphore, Sybil leverages computer vision, speech recognition and NLP to understand behavioural cues like the tone of voice, eye and facial movements or other non-verbal body languages. It then gathers the data to assess the person’s emotions, mood and attitude. It also generates an actual digital emotion scorecard with percentages and insights on the individual’s engagement.
Source: Sybil
The use of such platforms for sales calls is only the tip of the iceberg, and plentiful companies are providing such solutions. For instance, Balto not only analyses the emotional intelligence depending on the conversation audio but also shows agents the best things to say, scores 100% of calls, and alerts managers for coaching moments in real-time.
Source: Balto
While these solutions can be integrated with virtual meeting platforms like Zoom, Zoom itself has announced the introduction of Zoom IQ for Sales, which provides the meeting host with post-meeting conversation transcriptions and sentiment analysis. The application uses Zoom’s in-house automated speaker recognition and natural-language-understanding system integrated with Salesforce.
Intel+Classroom
Recently, Protocol reported Intel has collaborated with Classroom Technologies to leverage their solution of Class, a virtual school software integrated with AI and running on top of Zoom. The system can detect when students are bored, distracted or confused by assessing their facial expressions and how they’re interacting with educational content. Intel has come under huge criticism for this. Intel developed the analytics system based on videos of students in real-life classrooms and a team of psychologists who labelled the emotions and fed the data to an algorithmic model categorising student emotions.
These are just two main ways AI is seeping into our daily lives, some gaining criticism and some using. But similar AI solutions are everywhere. This includes delivery solutions for faster last-mile deliveries, passenger vehicles with driver monitoring AI techs such as Affectiva and Visteon, emotion-monitoring systems in Chinese prisons and businesses, or the several uses of Clearview AI and the likes.
Privacy concerns but growing companies
It’s safe to say, the controversies and disagreements toward such use-cases have only increased with the introduction of more applications. While these softwares are promising analysis only after all parties have consented to it, they still threaten the basic freedom of expression, or the freedom not to express, that any individual would have. Experts have also argued against the ability of the AI to rightly categorise the several complex facial, bodily and physiological responses humans may have. Furthermore, social media algorithms have often come under fire for tracking user activity and pushing content based on their life online. The privacy breach is taken to the next level with applications like these.
Regardless, the companies are growing. Uniphore recently raised $400 million in series E funding at a valuation of $2.5 billion, Class raised $105 million in a Series B funding, and Sybill has raised $1.6 million in a Pre-Seed round. Uniphore has also acquired three companies, their recent acquisition being Colabo, a start-up specialising in extracting and utilising information from structured and unstructured documents in real-time. While we are still to uncover the true risks and benefits of this technology, it is more vital than ever to ensure privacy checks, data protection and security frameworks to protect individual rights.