Working with CRED's Lending division, helping in strategising credit lending and collectionsKey metrics responsible for: NCL, Collection Efficiency, Cost saved per resolution Store Experiment – Concept and Execution* Carried out an experiment to collect user address through CRED store* Established the POC with a resulting into net profit of XX Lacs in POS collection Legal Resorts for Collections and Analytics* Worked on development legal solution for loan defaulters at a scale. Setup user targeting strategy* Automated workflow for notice issue, case filing and advocate calling Implementing conversational GenAI IVR bot for cost savings* Lead the project and was responsible for end to end flow development, execution and operations of IVR bot

Algorithm of Time-to-complete for Hypercharger* Designing and simulating TTC for hypercharger by predicting various conditions like battery temperature rise, SOC increase based on current supplied and ambient temperatureRange Prediction Algorithm* Designing and simulating vehicle range prediction based on user's past driving pattern across various driving modesMarket Research for upcoming 2W/4W EV* 2W Rating Model: created comprehensive 2W rating system to predict customer’s preference at various price buckets* Created a detailed report on 2W market and benchmarked projection of market growth* Ranked the list of features which impacts the product placement through regression modelCommand Centre Dashboard: Portal for Telematics data analysis* Worked on development of a near-real-time portal for telematics data analysis with ADX Kusto* Metrics Tracked: Speed, Acceleration, Gyroscope, Battery, Geo-location, Temperature, Voltage and 70 other

Recognition:* Awarded “Most Valuable Player – Tech” for the year 2021* Awarded “Swiggstar – Strive for Excellence” for OND quarter 2021Projects I took:Pricing Experimentation:* Conducted various pricing experiments (classic A/B exp.) to determine customers’ willingness to pay.* Presented the insights to various stakeholders and leadership team* Mitigated COVID-19 pandemic impact through Peak/Non-Peak analysis.* Metrics tracked: Revenue per lost order, Cost per delivery, Avg. Delivery Fee, Last Mile, Incremental orders etc.Swiggy Genie Potential Customer Tagging:* Worked on the development of a dynamic user tagging mechanism from Swiggy’s entire customer base.* Developed various metrics of customer behaviour, did hypothesis testing and quantified it.* Created machine learning models for classification and clustering; used synthetic control methods

* Prediction and presentation of future revenue, Pipeline conversion, EBIDTA and other data stories from ERP data sources.* Deployed scalable solutions using statistical methodologies and machine learning algorithms

Understood Business Working and Insights* Worked with a client which is one of the top Event Management Company in North America.* Gathered insights from the client's business and their databases.Understanding Concepts of Business Intelligence and Database management* Learned the concepts of ETL and Data warehousing.* Worked with SQL Server Management Services to extract data from EDW Databases from Dev Servers.* Implementation skills for Transact-SQL, DAX, and Query language MWorked with Visualization tools* Hands-on experience with a data visualization tool, Microsoft Power BI* Developed dashboards with revenue data for pre month and post-month analysis. Dashboards were integrated with the client’s custom-designed CRMWorked with Predictive Analysis techniques* Worked with classic predictive analysis methodologies for revenue forecasting* Considered various parameters like Pipeline, Pickup, and Revenue pace by 30 & 60 days before revenue, etc.* Designed time series models using ARIMA and Regression. Achieved 72% better prediction than sales’ team prediction

Worked as a Machine Learning Intern. I along with my other team members was assigned a project titled “Extracting Structured Data from Commercial Invoices and identity documents using Deep Learning”. This research project was aimed to extract generic data from already existing commercial invoices and store it in a database. As a part of the project, I designed the project layout, developed a model for text extraction, designed a flask based app for demonstration and worked on a deep learning model for document identification

I worked as a Machine Learning Intern. I was assigned a project of developing a Deep Learning model on the YOLO via Keras on Tensorflow framework. The model was expected to detect customers in a restaurant and generate a seat matrix. The end result gave an accuracy of over 93%.

I was assigned to High-Voltage Gas Insulated Switchgear Factory where I had my orientation training and was assigned a project of designing, developing and prototyping an acoustic meter at fraction of cost which intend to detect flashovers in GIS bays during its high voltage testing. This project was held under the Research and Development Department of the above stated factory.