Aug 2021 – Aug 2023 · 2 yrs · Gurugram, Haryana, India

• Search & Recommendations, CX Platforms
• Key Achievement: Personalized search results to the users by introducing User-Document level information in Learning-To-Rank model leveraging user metadata, document metadata and past interactions resulting in upto 50% more user-specific content and NDCG uplift by upto 2%
• Developed a Query to Key Initiatives classifier using fastText and used it to filter documents during search retrieval improving ranking precision
• Developed a Streamlit application that can simulate Search results given a query for a user and explain the LTR model reranking using SHAP
• Developed Entity Engine for Gartner Search comprising of entity exploration/detection, entity synonyms and entity coverage in search queries
• Integrated query-centric entities such as vendor, topic, industry from Entity Engine as features in LTR model, improving ranking quality
• Winning team in the Data Science Hackathon 2022 within Gartner given the challenge to understand client intent from search queries and map it to Gartner Taxonomy

Aug 2020 – Jul 2021 · 11 mos · Gurugram, Haryana, India

• Data Science and Automation, Services
• Key Achievement: Modeled a unified engagement metric at Gartner level based on logistic regression representing the value client received through various engagement channels. This is further used to measure quality of client calls and predicting their retention probability.
• Drove adoption of this metric through an automated Power BI dashboard giving actionable insights for call coaching service delivery associates.
• Conceptualized and developed “What-If” Scenario Analysis feature based on Pairwise Markov Networks which helped service delivery associates know ROI of their effort beforehand. This also provided a set of actionable next steps giving maximum client retention uplift.
• Implemented and fine-tuned Wav2vec model for call transcription to reduce WER. Further using the call transcriptions for downstream tasks such as keyword extraction, call summary, among others.

Jun 2019 – Jul 2020 · 1 yr 1 mo · Gurugram, Haryana, India

• Product Team, Peer Insights (Enterprise software and services reviews and ratings platform by Gartner)
• Key Achievement: Discovered, designed and developed automated personalized weekly emails containing market snapshot, leading vendors in industry, trending products, latest news and reviews readership digest to drive user engagement in the freemium product “Peer Insights Plus”
• Awarded for “Champion Mindset” title among 30 Rotational Associates for reducing time-to-market in product discovery experiments
• Contributed to platform growth at zero additional costs resulting in 8% share of reviews on Peer Insights through new channels
• Automated topic extraction from reviews using TfIdf, TextRank, Rake and led it to production as topic based filters in the platform with 4% CTR
• Setup end-to-end product metrics dashboard for continuous tracking of key metrics such as engagement, success, retention, cohort analysis

Jan 2019 – May 2019 · 4 mos · Kanpur Area, India

• ESC101A (Introduction to Programming)

May 2018 – Jul 2018 · 2 mos · Bengaluru, Karnataka, India · On-site

• UI Team, Ekart
• Implemented the app walkthrough in React Native for Ekart City Logistics android app for the first time users, hence reducing onboarding time
• Developed a functionality of screen recording in the app allowing users to report any issues via recording screen, facilitating easy debugging

May 2017 – Jul 2017 · 2 mos · Kanpur Area, India

• Remote internship under Prof. Siddharth Garg, NYU. Project aimed at crafting adversarial datasets for neural network architectures and designing networks for defense against those attacks.

May 2017 – Jul 2017 · 2 mos · Kanpur Area, India