SRM Deemed to be University bagged 1st place on Angelhack Globa
'Ankur Agarwal (first from right) and Krishna Maneesha Dendukuri from Next Tech Lab - SRM Deemed to be University'
SRM Deemed to be University Tech Lab Members bagged 1st place on Angelhack Global Hackathon 2018 at Microsoft Hyderabad
New Delhi, June 2018: On 3rd June 2018, a student team from SRM Deemed to be University, Kattankulathur bagged the 1st prize at Angelhack hackathon which was conducted as a part of Angelhack Global Hackathon series 2018 at the Microsoft campus in Hyderabad. The present year theme of “Seamless Technology” was set with the intention of giving developers the freedom to use any technology without restrictions to solve pressing problems concerning them and developers across the world. Solutions like these would enable quicker prototyping and deployment, increasing the already exponential rate of productivity of the technology industry.
Over 200 students, experienced Professionals and Entrepreneurs competed for this rare opportunity. After number of reviews and 2 rounds of extensive pitching sessions, the results were finally declared at the end of this 30 hrs long event. The Judging panel comprised of professionals from Microsoft, Amazon and major startups.
Two students; Ankur Agarwal and Krishna Maneesha Dendukuri from Next Tech Lab at the University, were a part of this winning team. The opportunity of attending the Global HACK celerator Program 2018 connected the students with thought-leaders and experienced entrepreneurs to help themselves in becoming more versatile while refining their ideas to build their hackathon winning prototype into a full-fledged start-up.
The team created an Artificial Intelligence based application which helps resolve bottlenecks for machine learning practitioners. Selecting the best algorithm to be applied on a given dataset in order to achieve the highest accuracy rates was the problem. This required intensive data analysis and a lot of trial & error cycles of applying the models while consuming a lot of time and processing power for training. This tedious process makes developers treat machine learning more like alchemy than an exact science. These students built a solution which can predict the efficiency of 10 different algorithms without even running them, for a dataset given by the user and therefore, suggest the best one to be applied by saving the valuable resources like training time and GPU power and bagged the 1st place.