Auto-translation used

From "Dungeons and Dragons" to AI-driven companies

I've always wanted to do something more than business. Create a project that will change people's lives right in the present tense, simplify some things or change our view of them.

Since 2017 , my partners and I have had a marketing agency, we have been engaged in digital marketing, provided consulting for sales departments, conducted training, and set up CRM systems. One of the training modules was quality control in the OP, that is, a person in the sales department had to listen to calls and identify weaknesses in the conversation of employees.

In the spring of 2024 the thought occurred to me: why would a human do this, if right now the neurons can accurately draw conclusions based on the text? The only question is how to turn sound into text, but it is also solvable.

Inspired , I called my classmate, who was engaged in artificial intelligence. I'm telling you, Kostya, there is an idea that when a call gets into CRM, it turns into text and is analyzed via ChatGPT. Is it difficult to do? And Kostya replies: "You won't believe it, we are just putting together such a project right now."

It's funny that at that moment Kostya and I were playing the NRI "Dungeons and Dragons" online. Who doesn't know – this is a tabletop role-playing game where a session lasts 4-5 hours. The whole process takes place within the framework of live communication between the participants and the game master, who comes up with the plot and plays the characters.

And so Kostya, for fun, wrote a transcriber that recorded and translated all 5 hours of the game into text. And then, with a request to the neural network, a summary of this call was made. As a result, he became the prototype of our current product.

We invested our own money in the development of the software and started trying to sell it. We started by communicating with clients, I conducted casting sessions, found out what problems there were in analyzing the work of sales departments, and we put it into the code.

The first few sales were done quite easily, I just told people that we could identify deviations from conversations – negativity or non-compliance with the script.

Two companies immediately paid 200,000 tenge each for this, although the product was still very raw.

We started implementing it, testing it on real cases, and in the process we realized: this is exactly the product I was looking for, it changes people's lives. We have the technology to create an AI assistant for anyone: sales manager, accountant, HR.

At this stage, we formulated the company's mission as follows: to save people from useless routine work and free up time and energy to realize their creative potential or perform more important, profitable tasks.

Next , I'll tell you how customers reacted to this idea and what our first experience with an investor brought us.

Comments 1

Login to leave a comment

Ваш рассказ вдохновляет! Особенно ценно, что уже на раннем этапе ваш продукт нашел отклик у клиентов, несмотря на его «сырость». Это говорит о том, что вы действительно решили актуальную проблему.

Reply