|
Listen to this story
|
MachineHack, in collaboration with Indus OS, successfully completed the ‘Quote to code I: Pandora’s box’ held from September 9 to October 8 for data engineers, data scientists, analytics professionals, and tech enthusiasts at all experience levels. The hackathon witnessed 600+ participants competing to develop various top-notch solutions.
Founded in 2013 by three IITians, and backed by PhonePe, Indus OS is building the largest ‘Made in India’ app store– Indus App Bazaar – an alternative Android app store. It offers users highly personalised and localised apps and is a gateway for app developers to distribute and grow their apps with maximum support and no restrictions.
The participants had to predict the top app recommendations from the metadata provided. Then, they performed multiple tasks and transformations to derive metrics to impress the judges.
Let’s look at the best solutions for the Indus OS hackathon & learn about the winners’ experience.
Rank 01
Arshdeep Singh, a senior data scientist at Vedantu, secured the first rank at the hackathon. For Singh, the hackathon was a fun and enriching experience. After completing his computer science degree from IIT Bombay in 2019, Singh joined Vedantu. He said he looks forward to more such hackathons on the MachineHack platform.
Check out the winner’s approach here.
Rank 02
Varun Jagannath bagged second place in the hackathon. The data scientist from Hyderabad currently works at DBS. Jagannath thinks of MachineHack as one of the most useful platforms for data science. He followed the retrieval, filtering, scoring, then ranking approach.
Check out the winner’s approach here.
Rank 03
Siddharth Singh has worked as a data scientist at Karmalife.AI since 2021. The Kaggle expert from Bangalore has been participating in hackathons for the past two years. MachineHack is Singh’s go-to platform for competing with some of the greatest minds in the domain.
Singh’s best single model is a 5KFold lightgbm. He said that the interesting part of this competition was that we had to generate, train, and test ourselves. The candidate generation strategy is the key to going beyond the limits of accuracy.
Singh’s solution is based on various retrieval strategies, feature engineering, and GBDT.
Check out the winner’s approach here.
Ending on a high note
The month-long hackathon saw some remarkable solutions from the participants. The development of these unique solutions was impressive and underscored the speed at which the community can develop new ideas and applications. Apart from cash prizes won by the top three winners mentioned above, the top scorers will get a chance to get hired by Indus OS for roles across seniority. To learn more about the hackathon and see all of the 2022 hacks, visit MachineHack.
