Search our Blogs
Showing results for 
Search instead for 
Do you mean 

AngelHack 2015 London - From developers to developers

Please note that HP IDOL OnDemand is now HPE Haven OnDemand. The API endpoints have changed to Haven OnDemand. Please see the API documentation for more details.




Seeing IDOL OnDemand in action over the weekend was a great experience for a first-time attendee at a hackathon, like me. I enjoyed seeing the developers’ positive energy and the joyful atmosphere despite the long working hours. They worked hard throughout the weekend in order to create the best possible application in the hope of being the triumphant team to represent London at the final round of AngelHack in Sillicon Valley.


picture1.jpg     picture2.jpg


Working hard despite the long hours …


The first thing that struck me was the diversity of the hackathon attendees: different age groups, nationalities and professional backgrounds. Throughout the weekend I have met people from all over the world, from college and university students to software developers, start-up founders and business professionals. I was pleasantly surprised to see that the event wasn’t dominated only by male attendees, but also by some female developers, one of whom was no older than 17 years old. Kudos to her for taking part in a hackathon at such an early age, and very well done for coming up with an idea and implementing it all by herself within 24hrs!


Our developer team was there to provide assistance not only in how IDOL OnDemand works, but also in explaining the concept of Big Data and how our product incorporates cutting-edge technologies. Our task throughout the weekend was to encourage developers to use our collection of APIs and incorporate it with their application in order to obtain smart analytics out of their data. Most of our end-users found it very easy to get started on using our APIs, even only with basic programming knowledge. We encouraged them to go on and freely explore our system and the functionality available before deciding which APIs might be useful and within a couple of hours we had more than 10 teams deciding that they want to base their application on our platform.




By interacting with so many different teams during the hackathon, one aspect that I found challenging but intriguing at the same time was to get a deeper understanding of their projects. I had to understand the problem they were planning to deal with, the solution they had in mind, and how they were planning to achieve it using IDOL OnDemand. Whereas some of them focused on obtaining smart analytics from Text Analysis APIs, others decided to use a combination of different APIs across the system varying from Speech Recognition to Sentiment Analysis and Predictive Analysis. Of particular interest was the Recognize Faces API, still in Preview mode and not yet available to the public, but which raised interest after the Nerf Gun demo included in our Keynote presentation. We have tackled a variety of topics from monitoring sentiment analysis in news articles, to tracking animal movements in their natural environment and trying to predict potential dangers. In all the projects we had to help them select the series of APIs that would provide the most useful behaviour, explore the multitude of parameter options available and even provide assistance in training the APIs with their custom datasets.


picture4.jpg     picture5.jpg


Nerf Gun Video:



Out of the 26 different projects that were submitted at the end of the hackathon, I can proudly say that more than half of them used the IDOL OnDemand APIs, including the overall winners of the hackathon. Five out of six top projects incorporated our technologies, and I must say that the ideas presented were original and innovative. The winner of the HP prize we offered was a mobile application that used the Concept Extraction and Sentiment Analysis APIs to extract sentiments from tweets regarding the London Underground in order to detect any delays on lines. Based on that, the app would set the alarm to wake you up in time to get to the 10 o’clock meeting you have scheduled. I have witnessed the evolution of the app throughout the whole 24 hours and I must admit that the development was carried out smoothly, with a clear architecture in mind and a vision of what the end result should look like. They needed minimal support in using our technologies and successfully integrated our APIs with their product.


The overall winner developed a web application called NewsSense, which makes sense of the news! Their app used Entity Extraction to obtain relevant information from the news about competitors on the market and then Sentiment Analysis to present the overall feeling about the researched company across media in real-time. What’s funny about the winning team is the fact that until 2am on Saturday night they hadn’t decided which idea they wanted to pursue and they kept changing their minds whenever we used to go around to check up on them. Despite this, they managed to get a fully functional piece of software in less than 12 hours! Impressive! Very well done to them for managing to develop a sophisticated piece of software with a shiny interface in just a few hours.




Winning team


Despite the long working hours and the fact that we managed to sleep only about 5 hours across the entire weekend, the hackathon was an interesting experience that offered us the insight and the motivation to work hard towards offering the best possible platform to our end-users. Coming in direct contact with the developers gave each of us a rewarding feeling for being able to help them achieve their goal. Well done to everyone in the team for volunteering their personal time and I look forward to the next one already. :)


picture7.jpg     picture8.jpg


IDOL OnDemand team

Social Media
About the Author
† The opinions expressed above are the personal opinions of the authors, not of HPE. By using this site, you accept the Terms of Use and Rules of Participation