Snorble solves the challenge of children’s speech recognition thanks to Sciling

Snorble is the brainchild of the serial entrepreneur and father of two small children, Mike Rizkalla. After years of trying to establish a healthy bedtime routine, he became convinced that his own son was, in his words, “an incredibly adorable descendant of Dracula”.


Sleepless nights, nightmares, inconsolable weeping, fear of the dark, of sleeping alone… Many parents swear that the months they have endured with hardly any sleep due to problems getting their children to sleep have been one of the most physically and mentally exhausting experiences of their lives. It’s not surprising that a lack of sleep is widely recognized as one of the most cruel methods of torture that exist.

Getting his children to sleep was almost a matter of survival, but when Mike tried to find a solution to the problem, he realized there were millions of answers available on the Internet, but nothing the really met his needs or those of other parents in the same situation. Convinced that “necessity is the mother of invention”, he decided to turn it into an opportunity to develop a product and re-think small children’s bedtime.

Never before had a product been developed like the one Mike had in mind. More than just a sleep trainer, it was to be a companion for children, a friend who turns bedtime into a magical experience. Creating it was going to be quite an experience. Its physical appearance alone involved coming up with 125 different concepts! But the challenge didn’t end there. Soon Mike realized that in order to give Snorble all of the functions he wanted, he would need a world-class team of experts in Artificial Intelligence.

About the client:

Snorble is a start-up from New York that develops technology and products based on Artificial Intelligence. Their star product is an interactive toy that can be customized and which helps children develop a proper nighttime routine and healthy sleeping habits.

The challenge

Building a digital assistant for children up to four years of age to gamify their daily routines and which would work using a small IoT device was not going to be easy. Even more importantly, if they were not able to prove that a conversational aid could be created to meet these demands, Snorble might not make it off the ground. So, after confirming the extensive experience in language recognition we can count on in Sciling, Mike didn’t hesitate to place this part of the development into our safe hands.

It is never an easy task to deconstruct oral language into its components and turn it into text to be processed by a machine, then extract the true intentions behind those words and construct a coherent response. What’s more, if it is a child speaking, the difficulty is even greater. So, being able to count on a team of academic doctors specializing in speech processing like the one we have in Sciling was essential to the success of this project.

“After many years at the helm of my own company of innovative IT solutions, I knew exactly what qualities would be needed in the team that was going to take charge of this part of the project. It was enough to read the initial report on risks they sent me to realize that Sciling was exactly what we needed. They have a systematic, meticulous way of working. They are not just another technology provider. They go the extra mile. They understand the business. After months of searching, I wasn’t able to find any others like them,” says Mike.

Strategic vision:

With such a technologically complex project, Sciling had to identify the main technical risks that would have to be tested as soon as possible in the proof of concept. This enabled Snorble to be the successful product it is today, achieving 100% of its target less than 24 hours after it was launched in the crowdfunding platform specializing in technology, Indiegogo.

Our solution

The result of the intensive collaboration between Snorble and Sciling is a conversation aid that delights both parents and children and which encourages kids to take up healthy sleeping habits. Snorble has a series of unique characteristics that make it stand out, ranging from addressing the child personally to find out why they are having trouble getting to sleep, to telling them personalized stories or helping them relax. And all of this is done with a modular design that enables new functions to be added quickly and simply in future.

Snorble has been driven by creativity as of the moment the idea was conceived, and with that in mind, its ecosystem will continue to evolve. Working with partners like Sciling, they plan to keep exploring a host of exciting projects in future.


Month for the first proof of concept


Months for the first prototype

In Sciling they are wizards. They are a team of doctors and engineers able to build incredible things using Artificial Intelligence.

Mike RizkallaCo-Founder & CEO

The robustness of Sciling’s proofs of concept is better than other suppliers’ final products.

Mike RizkallaCo-Founder & CEO

The implementation process

In order to know if the technology developed was going to meet the business needs or not, it was very important to understand how Snorble would work in a real environment. What’s more, in the case of a start-up, the work is always against the clock, so it was vital to find out what technical aspects definitely had to be tested in the proof of concept. “The smooth communication we had with the team at Sciling at all times and their ability to understand the business and identify its most relevant aspects enabled us to develop the first robust prototype in record time,” Mike enthuses.

Sciling is aware that it is essential for an entrepreneur like Mike to not only be able to validate an idea solidly and demonstrate it to others, but also to bring all the work into line with the business requirements so as to optimize the allocation of resources and capital. To do so, they ensured that the project’s technological risk was reduced to a minimum, including the most significant aspects in the proof of concept carried out. “Sciling not only helped to put my uncertainty to rest, but also to save time and money, which has allowed me to turn my idea into reality,” Mike asserts.

Technologies used

Synthetic Data Generation

To streamline the training of models

Automatic Speech Recognition (ASR)

To recognize specific speech

Natural Language Understanding (NLU)

To understand messages and intentions

Deep Learning

To get state-of-the-art results

Why us?

  • We have been helping start-ups for years to ensure the viability of their business models from the technological perspective.
  • The proofs of concept for quality that we carry out enable any business venture’s risks to be minimized.
  • Natural language processing technology is part of our core.
  • We can count on a team with over 15 years’ experience in Machine Learning.