Case Study: Societe Generale builds predictive models and scales its ML hackathon with HackerEarth

Building predictive models from banking and financial data

Societe Generale, through its Global Solutions Centre, wanted to put its massive banking and financial data to better use by crowdsourcing data analysis and building predictive models, and to scale its flagship Brainwaves event with a Machine Learning theme. To reach both top students and experienced developers and build mindshare in India’s data science community, Societe Generale partnered with HackerEarth and used the HackerEarth Data Science platform to run the challenge.

HackerEarth ran a 30-hour Machine Learning hackathon on its platform, providing a customizable ML challenge, management and validation of submissions, and a customized auto-evaluation mechanism (scoring on 50% of test data live to prevent overfitting, then full evaluation after the contest). The event drew 1,893 participants, 306 submissions (with over 50% experienced professionals and students from India’s top 10 engineering institutes), produced three winners and awarded $10,000 in prizes — demonstrating HackerEarth’s platform and managed service capability at scale.

Societe Generale

Rajesh Karuvat

Sr. Vice President