Vatsal Shah

- Leveraged AWS Comprehend to redact PII’s from more than 1.5M customer call transcripts and used Mistal 8x7B model in AWS Bedrock to classify the call transcripts into predefined categories, achieving a 93% accuracy. - Created a ML pipeline to predict loan charge-offs using XGBoost Classifier model through AWS Sagemaker and achieved an 85% accuracy, saving $2.25M for the company every year. - Engineered an end to end pipeline in AWS to summarize 25K customer survey responses we receive monthly using Claude 2.1 model to get top 5 actionable insights for different business units, saving over 75 hours of human effort every month. - Utilized Hugging Face embeddings to vectorize the monthly insights generated and performed K-Means Clustering to group similar insights, helping business prioritize and track the issues. - Collaborated with the Business Strategy team and provided insights on customer survey text responses by providing them with sentiment scores, bigrams, and word clouds using NLP.

- Analysed 250k customer surveys, identified the root cause of low customer satisfaction scores using SQL and data visualisations created in Python, and provided recommendations that would increase the overall satisfaction score by 5.4%. - Created ETL pipelines for 4 Tableau dashboards by fetching data from AWS Redshift and executing a Python script in AWS Glue, eliminating the need for manual updates and reducing processing time from 8 hours to 15 minutes. - Developed a Random Forest Classifier Machine Learning model to flag customer complaints, achieving a 92% accuracy, which led to enhanced customer loyalty.

- Conducted extensive exploratory data analysis and identified significant events for several industries by analyzing a large dataset of 1.6 million rows of transactional data, with a goal of providing recommendations as to which industry should they target. - Utilized time series forecasting techniques such as ARIMA and ETS to analyze trends and detect seasonality patterns in the data, measured accuracy using MAPE and derived a MAPE of as low as 4.5% for some industries with a 95% confidence interval. - Conducted in-depth research on 30+ financial indicators and developed a heat map to crucial factors that significantly impact these industries, and developed a comprehensive metric that will enable Bill.com to track the progress of these industries, based on the most significant indicators.

- Contributed to enhance 20 JIRA story creations, by delivering technical requirements, for the development team to make project planning efficient by 15% towards a client’s post sales service software. - Devised workflows and triggers as a Salesforce Administrator to improve Einstein classification models that resulted in increased efficiency in routing customer complaints to the right customer service employee by over 30%. - Built an automated workflow to send email notifications to case owners based on severity of complaints, resulting in a 25% reduction in complaint resolution time compared to the manual approach. - Collaborated with Quality Assessment and Release teams to deploy 3 projects to the production environment using Agile methodologies, receiving a special mention award for exceptional teamwork in one release.

- Conducted Extensive Market research of CRM companies for International Markets, successfully compiled a list of over 1500 companies from 8 different countries. - Reached out to Managements of CRM companies globally and collaborated with them on a revenue sharing basis, generated several leads and developed good relations for the company. - Presented our product to over 40 Managements of CRM companies and successfully landed 2 clients for the company.

- Worked in the Public policy Department. - Was the Liaison to several political dignitaries all over the country, on behalf of the organisation. - Organised MUN conferences in Italy, Delhi, Srinagar, Udaipur, Mumbai. - Was an integral part of the social movement organised by the organisation called “Plastic Hatao Desh Bachao”.

Relevant Coursework: Foundations of Statistics, Data and Programming, Machine Learning, Management Science, Forecasting, Fundamentals of Financial Analysis, Business Data Management

IETE-TSEC : Marketing Head- 2019-2020 - Headed the Core team of internal Smart India Hackathon. - Lead a team of 12 Marketing Executives to organise the fest Ellixir’19. IETE-TSEC : Marketing Executive- 2018-2019 Organised fests like ELAN’18 and Oscillation’19, and was successfully able to convert brands like Sony MIX, Society Tea, JAVA, Wok Express to come in as fest sponsors.

Participated in the Handball DSO tournament and achieved the third place.