Last year’s hackathon focused on smart city solutions and many of those projects. In light of that event, it came naturally to us to move forward with other innovative projects leveraging Artificial Intelligence as the main technology this time around. Our partners from Orange joined us for the second year in a row and together the participants developed some progressive apps with a real potential to go mainstream or at least become commercially viable in the not so distant future.

Over 70 specialists from all walks of life attended the hackathon. Both Tremend and Orange genuinely believe in skillset diversity and that’s why this year’s “br.A.I.n marathon” welcomed and included top developers, talented UX designers, highly experienced project managers, and visionary marketing and sales specialists. This mix of talent and human resources resulted in an unprecedented event in Romania. Now, are you curious to find out more what all the fuss was about?  

And the Hackathon Winners Were…

The 1st Place: Mind Your Code – MinDraw

The winning team focused on the social aspect of Artificial Intelligence, showcasing a solution for children with special needs suffering from autism or Down syndrome. The interactive application, which can be used in multiple languages to develop cognitive and people skills, analyzes a child’s drawing, figures out the drawn object and associates it with a real-life object. Leveraging technologies such as neural networks, Python, Django, TensorFlow and Keras, NumPy and Swift, the application has a high degree of feasibility and a solid business model that can soon make this solution a reality for many families and children around the world.

The 2nd Place: Mind Install – Yellsy

The runner-up team designed another social solution, presenting a mobile application that can identify a baby’s needs when crying and alert parents. The team used a data set as the cry translator for classification analysis. The most challenging part of the prototype was removing the background noise so the AI agent could correctly identify a baby’s emotion.

The 3rd Place: brAInard – Speech2Task

The third place winner team developed an integrated speech-to-text and translation prototype that can summarize conversations from meetings or audio recordings. Developed both for web and mobile, the solution powers Microservices, Cloud and API GW technologies and is intended for corporate as well as personal use.

Here is the Full List of Projects:

Every single participant gave their best and deserves our recognition for their extraordinary achievements. Here are all the other incredible prototypes, in no particular order:

Bito’BrAIn – The Sortifier

A shopping list management solution with a rather catchy and suggestive name, The Sortifier mobile app helps users manage all their expenses in one place by scanning (taking a photo) of their receipts. Employing Python for the backend side, neural networks (Random Forest) and a word-embedding linguistic model from Facebook (FastText), the app stands out through its ability of learning new words, data and product classification methods.  

M.I.M. – Shelf Management

The retail-centric app is an intelligent Python-developed solution for stock management in a supermarket. The AI agent, called SHAI, ensures a constant flow of products on shelves by analyzing and recognizing products and empty spaces. This is done through a planogram which is realized with video cameras aligned on the opposite shelves. It is another great example of automation that reduces human error and repetitive tasks.

UMA – My Travel Buddy

The travel industry is currently a tough market to capitalize on, given the multitude of options and apps. However, My Travel Buddy is an ambitious combination of Airbnb meets Samsung’s Bixby and stands out as an app that integrates e-Wallet, book flight & accommodation, trend anticipation, and user profiling features, among others. Leveraging Machine Learning, the mobile app brings travelers together by analyzing their behavior and anticipating their travel needs.  

Piece Of Your Mind – leCHEF

For the gourmands out there with a passion for cooking up recipe ideas out of forgotten food in the fridge, leCHEF is the perfect mobile app solution. The Python-developed app, which offers Facebook and Instagram integration, uses image recognition (via smartphone pictures) for food products while a bubbly AI bot assistant takes care of the communication part (text to speech, speech to text). Great way to stop food waste, try out new unique recipes and have fun all along!  

The Drunken Hackers – Senses CX

Marketing and sales automation as a SaaS solution is no longer something new under the Cloud! New solutions targeting this field must be niche-oriented to bring added value to corporations. And that’s exactly what Senses CX does — using AI Sentiment Analysis technology, the app provides valuable insights into a brand’s status and products as perceived by consumers. Another Python-based solution that can take the pulse of a company’s digital popularity!

Autonoma – Employee Retention App

The dream of every HR department! Studies on the costs of replacing an employee leaving the company keep producing various results — but these are never too optimistic! Employing HTTP API Service for the frontend and Python for the backend, the employee retention AI solution aims to anticipate satisfaction at workplace and assist HR specialists in developing intelligent solutions to reduce turnover.  

br.A.I.niacs – Attorney

A solution that integrates GDPR, road rules, human rights, construction rules, and other citizen rights into one single mobile app! With many tech combos, from UiPath community edition (for extracting law references) to Google API Translate API and Neo4J graph database (for AI-generated tags) to Google Natural Language, the app is intended for both corporate and personal use. An interactive and dynamic law-related solution with real-time updates and recommendations!

AI-Luminati – The AI in the sky

With an audacious name and a pitch presented as a stand-up comedy moment, the solution was by the far the most laugh-grabbing project of the hackathon! However, beyond the entertaining presentation, the infrastructure security app is extraordinary in its intended use. Leveraging Google Vision API and image content analysis, the app detects incidents (fires, floods, traffic jams, car accidents) within a matter of seconds and is intended for the city hall, police officers, as well as small businesses. A true AI in the sky!

Pinky and the AI – InterActive

When Tinder meets activity planning, you get InterActive, a user-friendly mobile app that lets you swipe right for your activities of interest and over time learns all about your preferences. Leveraging Machine Learning for consumer behavior, Universal Recommender (for personalized recommendation) and suggestions taken from Google Maps API, the app gets end-users out and about to experience fun activities relevant to their liking.   

Hackathon Prizes

The prizes for the winning teams were envisioned as an extended hackathon, with awesome team-building experiences. However, this time around, the winners get to brainstorm fun ideas in fun places:

  • A weekend in Prague for the team that won the grand prize
  • A 2-day ski trip in Romania for the team that came in second
  • A rafting trip in Romania for the team that won the third place

But the prizes didn’t stop there! All the participants and organizers were rewarded with a gift card at a popular airline, for a perfect trip… in the Clouds!

Lessons Learned

Here are some valuable tips and tricks straight from the source, aka the participants… since they know best!

  • Prepare and visualize your end goal! Try to shorten each step to build your project with least resistance. Less is more — do one job but make sure you do it well. Tudor Crîngu?, UX Designer at Orange, hackathon team — Masters of Intelligent Machines (M.I.M. Shelf Management)
  • Think big but start small! Instead of building a feature-packed application in a very limited amount of time with the risk of not finishing it, start by developing smaller features that will work in the end. Stefania Budulan, Software Engineer – Artificial Intelligence at Tremend, hackathon team — Bito’BrAIn (The Sortifier)
  • Don’t spend too much time on details! The idea is the most important and the way you present your solution. It is not critical to write a very clean code as it is to have a working solution within 24 hours. Lucian Gruia, Senior Software Development Engineer at Orange, hackathon team — UMA (My Travel Buddy)
  • Become one with the team! Your success or failure is a collaborative effort. Adrian Iova, Software Engineer at Tremend, hackathon team — Piece of Your Mind (leCHEF)

There is little doubt that this year’s hackathon was the most challenging of them all, given the complexity of each line of code. In fact, almost all prototypes left plenty of room for improvement and additional features, simply because powering fully-fledged AI-based applications within 24 hours was beyond human capabilities. The dynamics between the competing teams and the jury were on point and considerable brainstorming ideas related to AI technologies in the context of a business model were generated during the pitch presentations and Q&A.

Overall, it’s safe to argue that br.A.I.n Hack 2018 was an event full of HACKtion and we’ll get to see some real applications that will soon go into the funding stage and hopefully end up on our smartphones.

Which project did you like most or wish to find out more about? Contact us for more information.

Leave a Comment:

Your email address will not be published. Required fields are marked *