The key part of learning is feedback, not the transfer of knowledge. Especially since knowledge can now be easily found on the internet. But feedback is a problem 🙁 Because it’s too costly for teachers to provide it to each student in the needed volume. It was expensive! Until AI came along. Here are examples of how AI is already being used for feedback — and not even in traditional education.
THE ESSENCE OF THE PROJECT
Today’s startup was only founded in January of this year. They are still running pilot implementations, and there are not many details about the product on the website. However, even at this early stage, they have managed to attract the first 200,000 pounds sterling (about 270,000 US dollars) in investment.
TrainThis is a tool that automatically analyzes the communication of employees with company clients and partners, instantly providing feedback on the quality of how they do it.
To do this, TrainThis needs to be integrated with the company’s email platforms, voice calls, and video conferences – after which the AI engine of TrainThis begins to analyze all conversations, online meetings, and messages in these communication channels.
The employee himself receives instant feedback after each analysis conducted by TrainThis, which not only points out mistakes and shortcomings in communication but also praises for what was done correctly.
This is an important detail that many managers omit from the concept of “feedback”. Because you can’t scold all the time – initially, a person starts to feel like a fool, and then they become one 😉. Research claims that before you scold once, you need to praise six times.
The results of all TrainThis analyses are displayed to managers on a special dashboard. Using this information, they can understand who needs to be praised to perform even better and who needs to… no, not scold — but to train, and specifically what to train them in 😉.
The main feature of TrainThis is that it allows the company to further train the AI engine in both factual formulations and the communication style adopted by the company:
- By uploading examples they consider successful,
- Explicitly specifying a list of rules that must be followed in communications.
It’s important and cool that in this way you can further train the AI engine not only in how it analyzes employee communications – but also in the wording and style it uses to provide feedback to employees.
In their LinkedIn blog, the startup provides simple arithmetic: “If an employee receives feedback on an action they perform 50 times a week and improves their communication by at least 0.1% each time, then by the end of the quarter, the quality of their communication will almost double!”
What’s Interesting
A similar technology is used by the startup Oliv. Their primary pitch is “Clone your best salespeople,” not “give them feedback” as TrainThis does. Oliv raised $5 million in investment.
Another example of a different application of similar technology is the startup Zenarate. They created a purely educational tool—a simulator that can mimic customer requests and analyze how trainee salespeople and call center staff respond. Zenarate attracted $18 million.
Thus, we essentially see the same technology, but different startups use it to create different products. These products differ because companies initially view them as tools to solve different problems.
And this conclusion is essential, worth reiterating. A product isn’t what we make but what we sell. And people buy not what we make but what we sell to them. What we put forward as the primary offer is what they’ll focus on. How it’s built and what other purposes it can serve are details they will delve into after purchasing.
A critically important part of TrainThis is the ability to further train the platform’s AI engine based on specific examples from individual employees. After all, every company has its uniqueness, depending on what exactly it sells and how it believes it should be done.
Moreover, the mechanism of mutual employee training, where some people, having already learned something, teach others, has several strong advantages:
First, it’s a much more scalable model. The number of trainers specifically hired by the company no longer becomes a bottleneck, restricting the speed and scale of training.
Second, when you learn from colleagues, you realize they’re essentially like you, so you can become just as skilled. It’s not the same as learning from Olympic champions 😉.
Third, those who teach others better, more deeply, and more firmly grasp what they teach. It’s like the old university joke where a professor complains about a dim student: “Explained it to him for the third time. Now I’ve understood it myself” 😉.
A platform for such mutual training of company salespeople has been launched by the startup Upduo. They received excellent reviews from clients, including the well-known company Motorola, and raised $4 million in investment.
To some extent, TrainThis’s scheme of retraining based on the experience of successful employees is another variant of implementing the same model in corporate training with two advantages:
- The model becomes even more scalable because one employee trains the AI, which can then teach any number of employees.
- The model starts to work not only during the employee’s training period but also during their actual work after the training. Or even instead of it 😉
Where to run
The most important thing that struck me about TrainThis today is their focus on automatic and instant feedback.
The catch is that feedback, not knowledge transfer, is the key part of the learning process – whatever we teach others or learn ourselves. Anyone can now find information (knowledge) on almost any topic on the internet. And nowadays, people go to courses, not for knowledge, but for feedback.
And that’s usually the problem. Because recording a lecture or even giving it live to several dozens of students is quick and easy. But to give feedback to each student in the volume needed for them to genuinely understand it – that’s time-consuming and complex. Consequently, it’s expensive. Hence, people are willing to pay for saving time and effort on this.
The examples of startups mentioned in today’s review show that AI can be taught to provide quality feedback – adapted to a specific topic, specifics, and even the style of feedback delivery.
Therefore, the general direction is the creation of AI platforms for feedback delivery, capable of adapting to the subject of training and feedback specifics. Such platforms can be applied in various forms in corporate training, online education, and self-study.
Considering that feedback is a crucial part of education, sooner or later, no explicit or implicit educational service can do without this component, which significantly saves teachers’ time and enhances the quality of education.
Thus, the potential market for such platforms is vast. The whole question is, where is it best to head with this right now? Where can you start making more money on this more quickly right now?
Where would you first go with a feedback platform?
About the Company
TrainThis
Website: trainthis.ai
Last round: £200K, 28.09.2023
Total investments: £200K, rounds: 1.