CONVERSATIONAL AI CHECKLIST FOR HR MANAGERS
10 Questions to Ask Your Simulation Training Provider
CONVERSATIONAL AI CHECKLIST FOR HR MANAGERS
10 Questions to Ask Your Simulation Training Provider
Conversational AI trains employees to communicate effectively through authentic interactions with customers, coworkers, managers, and others in realistic simulations. The technology mingles automated speech recognition with natural language processing and machine learning so employees can speak naturally in simulations just as they would in the real world. All the while, the technology is assessing skills and personalizing training. The benefits are many, like anytime/anywhere training, personalized learning that’s scalable, real-world experience without the risks, confidence-building, and more.
But not all conversational AI solutions are equal. See below for ten crucial questions to ask before selecting a conversational AI partner for employee training.
DOES THE LEARNER HAVE TO FOLLOW A SCRIPT?
Unlike conversational AI used in a chatbot where an answer always follows a question, conversations in simulations are free-flowing and interactive. Yet some conversational AI providers adhere to strict script following in simulations, making it difficult for learners to develop flexible skills. In the real world, employees don’t follow a script. For authentic simulation training, conversational AI should incorporate enough logic to accommodate interactions outside a script as long as the learner stays on task.
HOW ARE LEARNER ERRORS CORRECTED?
Some conversational AI technologies stop the flow of a simulation to provide instant correction. Constant interruption for correction could frustrate some learners and hurt the confidence of others. Natural language processing should be robust enough for the avatar to continue the conversation even if the learner makes a mistake while simultaneously evaluating the learner’s performance. Correction can either be built into the conversation, or personalized practice exercises can be provided after the simulation, and then the simulation tried again.
Do learners have the ability to role-play in the simulation?
In role-playing, the employee interacts with an avatar that emulates a real person in mannerisms and body language. Role-playing is the most authentic way to implement conversational AI, and is a necessity when reading body cues is essential to training. Avatars, or pedagogical agents, are a separate technology that conversational AI providers may or may not employ as part of their solution.
Is “learning by teaching” supported?
Conversational AI is a learn-by-doing technology. But studies show that when teaching, learners work harder to understand the material, recall it more accurately, and apply it more effectively. During simulations, the conversational AI should periodically ask the learner probing questions that the learner has to answer in their own words. This rounds out “learning by doing” and maximizes critical thinking development.
Can the simulation detect the demeanor of the trainee?
Understanding the tone of a trainee in a simulated interaction is often as important as what the trainee says. Does an employee’s tone reflect curtness, impatience, and uncertainty or rather helpfulness, empathy, and confidence? While commercially-available conversational AI technology cannot yet detect emotions through tone or sound, it does have the ability to decipher emotion through the word choice of the learner. This is called sentiment analysis and should be included for training scenarios that require sensitivity from the employee.
Do you teach fundamental skills to trainees before the simulation?
A trainee may need product, company, or situational knowledge before they enter a simulation. Some simulation providers can leverage their AI technology platforms to teach fundamental skills in advance through automated online sessions while other providers may specialize in just the simulations.
Is specialized equipment required?
Conversational AI can be just as effective on a desktop or laptop as it is in a virtual or augmented reality. Specialized equipment should be an option, not a requirement. Conversational AI vs VR in Training >
Can our organization edit training products directly?
Most conversational AI providers will provide a turnkey solution where employees simply log in online and start training. And for many organizations, this is sufficient, even preferred. But if an organization is accustomed to building and supporting its products in-house, the ability to modify, or build, the training curriculum directly may be desired. Authoring tools for AI are complex systems. If a provider states that you can build simulations directly, it’s often not something you do 100% on your own through an application but rather a process that takes you through custom development with them. Make a list of what you want to edit directly, and have a conversation with your provider on the steps required to make that happen.
What kind of reporting and analytics is provided?
A dashboard should provide easy access to individual and collective performance data. Also, recordings of conversations or role-plays should be made available to listen to the trainee’s audio and assess tone of voice. Behind the scenes, the raw data the AI collects from simulations should be accessible. This is used to train the machine learning algorithms, identify collective gaps in learning, and demonstrate the impact of learning on organizational goals.
