Responsibilities• Leading a Machine Learning team• Performed tasks which stretched from beginning to end of ML, such as Python Environment Setup, Data Analysis, Feature Engineering, Models building, Models testing, Script writing, MLOps• Machine Learning Pipelining• Manage multiple projectsProjects• Column Detection with text coordinates in Multi Column Pages• Documents Layout Detection• Address extraction with NER• Text Blocks detection• Information extraction from documents1. Key sentence extraction - finding a couple of relevant sentences in a document containing hundreds of pages (extremely biased unstructured dataset)2. Value Pair extraction3. Keywords extractionUnderstandings• Document structures• Tabular structures• Sentences structures• Noun Phrases structures

Production Architecture Planning• Worked in a team to plan the Architecture of an AI software solution keeping in mind the accuracy, scalability, feasibility and reliability of the product. Designed following aspects of the product:• AI Pipeline Modelling• Front-end flow• OCR selection: R&D on various OCR tools such as Textract, Tesseract, Nuance Omnipage, Abbyy Signature Detection on Documents• Created an AI model to extract signatures from documents with an accuracy of 91% and f1 score of 0.91 .Document Classification• Using Universal sentence encoder and Scikit-learn, I built a document classification model with an accuracy of 98%.

Lead Data Analyst: Lead Data Analyst for Web analytics. Develop and implement data collection systems and other strategies that optimize statistical efficiency and data quality. Created and implemented the guidelines for web tracking implementation. Work closely with management to prioritize business and information needs.Data Analysis:I’ve analyzed stepwise conversion process of website from User landing till purchase and provided suggestions for improvement. As a result the conversion rate (CR) increased by nearly 50%.I also analyzed the poorly performing Mobile and Tablet Devices. Close examination discovered many UI issues. These fixes increased the CR of mobile by about 30% and stabilized the CR of Tablet which was very fluctuating earlier. Reports Creation and Automation: Automated the Daily MIS and Daily Marketing MIS report by accumulating data from MySQL, GA, Facebook Dashboard, AdWords, Criteo etc. Created real-time hourly conversion report using google analytics (API) spreadsheet. Now we can monitor the health of every step during conversion and diagnose immediately should an issue surface. Designed the Web Tracking Model:Planned the entire URL tracking, event tracking model when the new website of Nykaa was produced which very effectively helped us assess various features and make improvements in conversion rate, search, filter usage etc. Google Analytics (GA) / Omniture / Mixpanel Admin: Implemented GA URL, event, ecommerce, search tracking etc., I’ve also performed the linking of AdWords and GA. I coordinated with the Adobe Expert and our developer team to successfully implement Adobe Analytics in both website and App, each in less than a month. Linked AdWords with Omniture. Now we can track all the auto tagged AdWords Ads in both GA and Omniture. Analytics Trainer: Provide Google Analytics and Omniture Training to new recruits, Marketing team and Brand team.