Gauri Chaware Email & Phone Number

United States

Charlotte, North Carolina, US

• Enhanced Audit and Compliance dashboard design by providing detailed transaction summaries using GPT-4
• Improved risk detection in 4 compliance areas using weak supervised learning & snorkel modelling, achieving 95% accuracy
• Built a novel algorithm that generates a risk score based on various factors, allowing stakeholders to prioritize riskiest businesses first, achieved an accuracy of 92%

Charlotte, North Carolina, US

• Productionized a classifier using MLFlow to prioritize price sensitive customers to increase revenue

Amherst, Massachusetts, US

Worked as a Teaching Assistant for CMPSCI 445 (Information Systems) for 45 students

Charlotte, North Carolina, US

• Improved the accuracy of Honeywell’s product pricing algorithm by 20%
• Automated the entire data extraction process of the pricing algorithm of Honeywell using SQL in Dremio, achieving a speed up of 90%

Amherst, Massachusetts, US

• Improved latent topic identification from customer restaurant reviews by training additional auxiliary tasks (NLI, NLI-STSB, QA, Paraphrasing, MPNet) on top of the generated topic embeddings and analyzed their quality.
• Improved Turney & BiRD scores by 10% from the base model performance

London, England, GB

• Pioneered a banking product recommender, mined frequent item sets & association rules, generated $500k in sales
• Boosted name matching process for customer leads using soundex & Logistic Regression, achieved 30% speed up
• Deployed a credit scoring model using logistic regression to reduce loan defaulters, reduced risk of default by 20%
• Analyzed customers adapting to digital banking via dashboards & automated data pipelines in Python & Tableau

London, GB

• Built an unsupervised ML topic model and word cloud in Python to - 1. Drive the probable management strategic decisions 2. Streamline the existing manual reporting process 3. Help Relationship Managers (RM) to assess.
• Established new trade prospects for 34 countries using ML and statistical models in R - 1. Analysed critical data elements like account behaviour, average balances, FX 2. Studied historical behaviour of existing.

London, GB

• Built forecasting models like ARIMA, STL, BSTS and ensemble models for over 10 countries to estimate the total volume of transactions expected by HSBC in the next year 1. Implemented stationarity and seasonality checks.
• Automated manual process of bill checking using Python, increased the efficiency by 40%, reduced 400+ man hrs/yr, won “Team Star” award for making meaningful contributions to the stakeholders, out of 600 employees
• Streamlined 3 financial crime analysis reports & helped in taking strategic decisions by giving business insights
• Reduced falsely reported suspicious transactions by 30% using text mining techniques & word embeddings
• Provided monthly insights to strengthen management information (MI) relating to financial crime risk globally for all the markets using Excel, VBA, SAS and Alteryx - 1. Generated a host of metrics to evaluate the.

DE

1. Configured private servers for expansion of data storage2. Detected faulty vehicular components by building an image classifier in TensorFlow, achieving accuracy of 96%3. Initialization and installation of Virtual Machines on the servers4. Configuration and deployment of "MANCATS" software integrated with vehicles

Master'S Degree, Computer Science (Data Science Concentration)

Bachelor Of Engineering - Be, Computer Engineering