Harshita Moondra

At Flipkart Internet Pvt. Ltd., I scale recommendation services for Big Billion Days traffic while keeping user-path latencies below a second. I’ve also optimized recommendation pipelines to reduce processing time and compute cost by 20–30%.

I designed and productionised recommendation capabilities across feed personalisation, in-session adaptive recommendations, trending recommendations, and Complete the Look. I integrated an LLM-based exploration workflow into the New to User Explore pipeline, contributing to a 1.3% increase in feed user engagement.

At Goldman Sachs, I built end-to-end regulatory reporting tools that automated data sourcing, validation, and processing, reducing reporting errors by 25%. I also developed a reusable data validation tool and a data querying tool that increased data retrieval efficiency by 30%.

My work spans Java and Scala backend systems, distributed services, and data-intensive applications. I’ve mentored junior engineers and interns through code reviews, design discussions, and onboarding.