Sowmya Kothakapu
Frisco
Summary
Machine Learning Engineer specializing in real-time fraud detection and risk analysis. Expertise in Python and AWS, with a proven track record of developing efficient machine learning pipelines. Strong collaboration skills with cross-functional teams to implement innovative security solutions that reduce false positives.
Overview
6
6
years of professional experience
1
1
Certification
Work History
Machine Learning Engineer
Comerica Bank
04.2024 - Current
- As a Machine Learning Engineer, I worked on developing a machine learning system to identify potentially fraudulent transactions in near real time. The objective was to reduce financial losses while minimizing false positives that could impact genuine customers.
- Designed and deployed machine learning models for risk analysis, customer behavior prediction, and fraud detection, enhancing decision-making processes
- Building end-to-end ML pipelines including data ingestion, feature engineering, model training, validation, and deployment using AWS SageMaker
- Developed generative AI solutions for internal knowledge search using LLMs with Retrieval-Augmented Generation (RAG), improving information retrieval efficiency
- Executed Agentic AI workflows to enable AI agents to autonomously retrieve data and reason through multi-step tasks.
- Implemented model monitoring and performance tracking to detect drift, ensuring sustained reliability and accuracy of deployed models
- Collaborating with data engineers and business stakeholders to productionize ML solutions aligned with regulatory and security requirements
- Environmental Skills: Python, MySQL, MongoDB, AWS, Azure, RESTful APIs, Jenkins, Docker, and Git, SageMaker, S3, Lambda, CloudWatch, IAM, CI/CD pipelines
PROGRAMMER ANALYST
American Express
08.2023 - 12.2023
- The project involved the development and enhancement of a centralized data integration and reporting system to support real-time business analytics and performance monitoring. The goal was to extract, transform, and load data from multiple internal systems (ERP, CRM, and third-party APIs) into a unified data warehouse to enable efficient reporting and decision-making for stakeholders across finance, sales, and operations departments.
- Analyzed business requirements and collaborated with cross-functional teams to create technical specifications that aligned with stakeholder needs.
- Designed and developed ETL workflows using [Informatica/Talend/Azure Data Factory] to consolidate data from disparate sources.
- Developed and optimized SQL queries, stored procedures, and views for data processing and reporting.
- Developed interactive dashboards and reports in [Tableau/Power BI] to visualize key performance metrics, enhancing data-driven decision-making.
- Worked with cloud platforms (AWS/Azure) for data storage and deployment pipelines.
- Environmental Skills: SQL, Python, ETL Tools ([tool name]), Tableau/Power BI, Azure/AWS, Git, Agile Methodology
Internship: Machine Learning
Goal Street
06.2020 - 12.2021
- During my internship at Goal Street, I worked on developing a machine learning-based credit card fraud detection system. The objective of this project was to identify fraudulent transactions in real-time by analyzing transaction data patterns. By leveraging various machine learning algorithms, we aimed to reduce false positives and enhance the security of credit card transactions.
- Data Collection and Preprocessing: transactions. Collected and preprocessed transaction datasets, which included both legitimate and fraudulent. Ensured data quality through techniques such as data cleaning, normalization, and handling missing values.
- Exploratory Data Analysis (EDA): Conducted exploratory data analysis to identify trends, patterns, and anomalies in the data. Visualized key metrics to understand transaction behaviors and gain insights into fraud characteristics.
- Model Development: Implemented various machine learning algorithms, including Logistic Regression, Decision Trees, Random Forests, and Support Vector Machines (SVM). Evaluated model performance using metrics such as accuracy, precision, recall, and ROC-AUC.
- Feature Engineering: Developed new features from existing data, such as transaction frequency and average transaction amount, to improve model predictive power. Utilized techniques like one-hot encoding for categorical variables.
- Conducted model evaluation through cross-validation and hyperparameter tuning using Grid Search, optimizing model performance and selecting the best-performing model based on comparative analysis.
- Collaborated with team members to discuss project progress and insights. Presented reports detailing findings, model performance, and actionable recommendations to stakeholders, ensuring clarity for non-technical audiences.
- Assisted in integrating the developed model into the existing transaction processing system, supporting deployment and real-time monitoring for effective fraud detection.
- Environmental Skills: Python, Data Cleaning and Preprocessing, Exploratory Data Analysis (EDA), Feature Engineering, Model Selection and Evaluation, Statistical Analysis, Algorithm Optimization, Data Visualization, Collaboration and Communication and Version Control.
Education
Master of Science - Computer and Information Science
SOUTHERN ARKANSAS UNIVERSITY
Magnolia, AR12.2023
Bachelor of Science - Computer Science
JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY
Telangana,India07.2021
Skills
- Data Science
- Data Visualization
- Python
- RESTful APIs
- MySQL
- MongoDB
- PostgreSQL
- AWS
- Google Cloud Platform
- GCP
- Azure/AWS
- Agile Methodologies
- Version Control
- Git
- GitHub
- Development environments
- Machine learning/AI
- Testing and debugging
- Analytical thinking
Certification
https://coursera.org/share/5e50e1c8c401dd160b5bf0fe432c784c
Timeline
Machine Learning Engineer
Comerica Bank
04.2024 - Current
PROGRAMMER ANALYST
American Express
08.2023 - 12.2023
Internship: Machine Learning
Goal Street
06.2020 - 12.2021
Master of Science - Computer and Information Science
SOUTHERN ARKANSAS UNIVERSITY
Bachelor of Science - Computer Science
JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY