A bit of history

Ever since I was a child, I've been passionate about knowledge. There was something magical about adding new information to your mind, mixing it with what was there before, and reflecting. It changed you. Forever.

My first encounter was with paper books. Then came the ebooks. Years later, the wonderful podcasts appeared, which radically changed my way of learning because I could do it while walking.

I remember that time as my own Minister of Education, asking myself every day: what do I want to learn today?

2023: A frenetic year

2023 was the year of generative AI. I was excited about the number of new projects that focused on my passion: knowledge.

Summarizing documents, extracting useful information, and even being able to chat with your texts.

But I always had a special fondness for YouTube. That infinite ocean full of information. Although very difficult to exploit. For me, YouTube was the largest library of podcasts.

Why did we create Askthevideo?

At that moment I knew: there was a need to organize that immense amount of information and help people understand it (learn, reflect, and ultimately transform).

AI would have to be a tool that freed us from the noise around us, and allowed us to perform tasks of true human beings: think and reflect.

I was incredibly lucky to come across a group of people who, like me, embraced uncertainty. So together, we decided to create the best virtual assistant possible. One designed not only to help users understand the content of any YouTube video but to profoundly transform that experience.

That was our clear and sincere objective from minute one. We could not (nor wanted to) aspire to less.

And now, where do we start?

That's exactly what I asked myself. We had a clear idea of what, but not yet the how.

We started with the basics: organizing the important information from the video. So, we decided to extract the key points and organize them by topics. Simple and useful.

To ensure our assistant was complete, we allowed it to expand on any point of information or answer questions.

And so, with touches of care here, and strokes of technology there, we managed to create a personalized tutor.

Yes, okay, but what technologies do you use?

I could tell you with very fancy words, like we use advanced RAG (Retrieval Augmented Generation) techniques, where we combine Natural Language Processing techniques and semantic searches in vector databases. All this to extract precise information from the video.

And then, we pass that information to a LLM (Large Language Model), which takes care of answering your questions in a clear and simple manner (thanks to prompts specifically created for that task).

But ... too much jargon. Just know that every time you ask the assistant something, you are using the full power of AI to get the best possible answer. We guarantee it.