Aug 2018 – May 2019 · 9 mos · United States

Apr 2018 – Jul 2018 · 3 mos · Bengaluru, Karnataka, India

• Sentient - Analyze various multilingual customer touchpoints and aggregate all data sources to predict whether a customer interaction will lead to an escalation
• Drove the project from conceptualization to execution as a part of Myntra Innovation Challenge '18
• Improved the precision by 4% by using TFIDF weighted Word2Vec embeddings to classify customer call transcripts and emails as escalations resulting in ROC AUC of 0.96 for the ensemble model

Jul 2016 – Apr 2018 · 1 yr 9 mos · Bengaluru, Karnataka, India

• SABRE - Smart AI Based Returns - Determining the probability of genuine return and green-channeling (pre-approve) it for pickup given a customers' purchase and return history
• Productionized the model enabling instant doorstep refunds across 100% of Myntra returns since June '17
• Deployed an XGBoost model (V2) having a ROC AUC of 0.82 compared to V1 with 0.67 resulting in a 13% increase in green-channeling and a drop in return on-hold (ROH) cases (38 point difference in return NPS) from 1.7% to 0.2%
• Won the ‘Effective Tech Implementation’ of the year award at IFF ’18
• Buyer Program - Inducing repeat purchase behavior using personalized notifications and feedcards
• Achieved significantly higher open rates (6.5% - almost 2X) for personalized notification compared to generic (3.5%)
• Increased browse to buy conversion from 6% to 10% resulting in 2% incremental revenue

Jan 2016 – Jul 2016 · 6 mos · Bengaluru, Karnataka, India

• User Segmentation (5 star customer rating)
• Clustered users into segments using RFM scores
• Direct impact on customer experience - Loyal customers offered premium services

Jun 2015 – Jul 2015 · 1 mo · Greater Bengaluru Area

• Architected scalable REST APIs utilizing the MEAN stack supporting a concurrency of 10K requests with an average response time of 490 ms
• MongoDB | MongooseJS | ExpressJS | NodeJS | Apache Benchmark | SoapUI