In this blog post, we will share the remarkable journey of an OpenCV University student – José Benitez Genes.
José is originally from Paraguay. He grew up in poverty and had to drop out of college due to financial hardships. Things looked bleak as he tried to run a small business to make ends meet.
“It was a mess,” he told me. But there was one thing going for José; his friends knew him as a hardworking and technically brilliant engineer.
It is the darkest hour, right before dawn.
Fortune Favors the Prepared
José got an unexpected invitation from two entrepreneurs – Nicolas Parziale and Tomás Manzitti, based out of Argentina. They wanted to build the technology to convert display refrigerators into autonomous vending machines and learned about José through a friend.
José took the offer. With a big dream and just $10 in his pocket, this ambitious young man landed in Argentina.
The team formed a company called Intuitivo with a bold dream to build the Infrastructure for the future of retail.
Product Idea & Preparation
Vending machines are expensive. Display refrigerators are cheap. Turning a display refrigerator into a vending machine had a massive appeal in price-sensitive Latin America.
One can solve this problem using different kinds of sensors (RFID, weight sensors, etc.) in several ways. However, installing, calibrating, and maintaining these sensors is an expensive operation.
José decided to opt for a computer vision solution because cameras and compute are cheap. A robust AI solution would not require calibration, and any untrained person could do the installation.
José embarked on a journey toward mastery of computer vision and AI. He enrolled in our courses and started building his dream solution. In his own words,
“I took OpenCV University courses in 2021. They played a significant role in the choice of algorithms and the quality of the Autonomous Point of Purchase (A-PoP) product we built at Intuitivo, and we were able to raise $7.8M in 2022. I highly recommend OpenCV University courses for entrepreneurs, students, and professionals who want to start a career in AI.”
The Product: Autonomous Point of Purchase (A-POP)
José’s solution was a three-camera system connected to a compute module with a smartphone app serving as the display interface.
They called it an Autonomous Point of Purchase or simply A-POP.
The user starts by scanning the QR code on the A-POP. The user is verified, and the A-POP door opens. The user picks the products they want to purchase while the cameras capture the interaction and send it to the cloud. The products picked by the user are automatically identified, and the correct price is deducted from their account.
It’s that easy. Here is a video showing Intuitivo in action.
The AI Backend
The ease of use for the user is accomplished through a sophisticated AI and MLOps backend.
A-POPs can hold any product inside it, and every product needs to be identified precisely with a high degree of accuracy. There can be thousands of unique products (a.k.a Stock-Keeping Units or SKUs). The AI backend is trained on just a few example images of the product initially, but the quality of the model is continuously and autonomously improved using Active Learning with each new purchase.
After every transaction, the data is curated by the Data Operation team, which uses internal custom software for fast curation. This data is fed into the Active Learning processes that automatically train and update their machine learning models.
Intuitivo routinely updates its backend with the latest technology. For example, recently, they incorporated the Segment Anything Model (SAM) fine-tuned on their custom dataset in their backend. Their fine-tuned SAM processes images 62 times faster than human annotators.
The company grew through the pandemic and already has an impressive list of clients, including Arca Continental and Nespresso. In 2022, Intuitivo raised $7.8M to expand its operations. Hundreds of A-POPs are operational at many locations in North and South America. The team has expanded to 35 people and they are planning to grow the team in 2023.
Top 35 Innovators from Latin America
In 2022, MIT Technology Review selected José as one of the top 35 innovators from Latin America. What a stunning achievement and reversal in fortune for this young man!
Dream Big. Start Small. Act Now.
José’s success has shown us that AI is rapidly sweeping every aspect of our life, and the winners will be those who take the initiative to get an education in AI and start working on their dreams.
If you are looking for a career in AI, it’s time for you to start now. OpenCV University provides programs and courses that build a solid foundation in AI and prepare you for a fulfilling career.