Yachna Hasija
@yachnahasija
Data Scientist at Target Corporation
Bengaluru, India
Data Scientist with 4 years of experience, adept at analyzing complex problems, extracting meaningful insights from data, and developing innovative solutions. Proven ability to build and deploy models that enhance business performance and drive growth. Passionate about leveraging data science to solve challenging problems and deliver impactful results.
Experience
Data Scientist
Target Corporation
Developed a Site taxonomy using GenAI with 75% accuracy that improved overall search experience for Target. Developed a hybrid model to recommend the top-5 complementary items for each seed item, resulting in a 7% improvement in Precision@8, and a 12% boost in attributable demand compared to an existing rule-based model. Created a review tagging model to tag guest reviews on PDP pages, which improved the F1 score by 22% and a substantial 30% average increase in tag coverage per review. Built an XGB-based model to predict the Enterprise Next Basket Size over 30 and 45 days, and executed an A/B test, which led to a 30% reduction in MAE and a 23% increase in Incremental Sales compared to the baseline. Implemented an Always-On campaign strategy to continuously engage the audience cohorts, and effectively minimize their overall churn rate and improve their GLTV score. Worked on various Storewide and Category-Specific marketing campaigns. Devised an Incremental Sales KPI to gauge the efficacy of marketing and sales promotion campaigns.
Data Scientist
GoMechanic
Assisted in digitizing spare invoices and integrating AI into spare parts processes. Leveraged CRNN methodology and trained a custom OCR model on 10K+ Text Corpus to extract Handwritten text with 80% precision. Utilized YOLOv5 Object Detection model to identify relevant classes from retailers’ spare bill books, achieving a 97% accuracy. Implemented cutting-edge visual and language-based multi-modal techniques to process PDF documents, enhancing document understanding for various types of insurance policies and reducing human effort by up to 80%. Developed a Few-shot classifier to improve performance on unseen and unlabeled classes. Collaborated with Engineers and Front-End Developers to integrate an Insurance Policy document feature into the Retail app. Built a customer support chatbot using sentiment understanding and NLP methodology to increase customer retention rate.
Education
Great Lakes Institute of Management
Post graduation program
Data Science and Engineering
University of Delhi
Bachelor of Science
Mathematics
Awarded with a gold medal and certificate of merit for securing first position in three consecutive years (2016-2019) of B.Sc. Mathematics (Honors).