Machine Learning Data Scientist

Yaswanth Reddy Yarrabandla

Data Scientist

Chicago, IL +1 872- 242- 8299 ************@*****.*** LinkedIn SUMMARY:

Experienced Data Scientist with 3+ years of expertise in designing, developing, and deploying machine learning and deep learning models to drive data-driven decision-making and business insights. Skilled in predictive analytics, NLP, time-series forecasting, and collaborative filtering, with hands-on experience in building end-to-end data pipelines, optimizing ETL workflows, and automating real-time analytics. Proficient in leveraging Python, Azure Machine Learning, AWS SageMaker, Hugging Face, LangChain, Power BI, and Tableau to deliver scalable solutions and actionable insights. Adept at collaborating with cross-functional teams, translating complex business requirements into innovative data solutions, and enhancing model performance to support strategic goals. SKILLS:

Programming & Scripting: Python, R, SQL, PySpark, NumPy, Pandas, Scikit-learn, Matplotlib, Seaborn Machine Learning & AI: Supervised & Unsupervised Learning, Random Forest, XGBoost, Logistic Regression, LSTM, Transformers, Predictive Modeling, Collaborative Filtering, Sentiment Analysis, A/B Testing Deep Learning & NLP: Hugging Face Transformers, BERT, GPT-based Models, Text Embeddings, LangChain, Feature Engineering, Model Fine-tuning

Cloud Platforms: Azure Machine Learning, AWS SageMaker, AWS Glue, AWS Lambda, S3, Docker, Model Deployment, CI/CD for ML Pipelines

Data Engineering: ETL Pipeline Development, Real-time Data Processing, Data Wrangling, Data Cleaning, Data Integration, Data Transformation

Visualization & BI Tools: Power BI, Tableau, Plotly, Dash, KPI Dashboards, Drill-down Reporting, Interactive Visualizations Statistical Analysis : Hypothesis Testing, Regression Analysis, Probability Distributions, Forecasting, Time Series Analysis Collaboration & Workflow: JIRA, Git, GitHub, Agile Methodologies, Cross-functional Team Collaboration, Model Monitoring & Optimization

EXPERIENCE:

Northern Trust, USA Data Scientist Jan 2025 – Present

• Architected and developed an integrated behavioral bias detection and liquidity stress simulation system by leveraging Azure Machine Learning, Python, PyTorch, LangChain, and Hugging Face transformers to analyze structured and unstructured financial data, enabling portfolio managers to identify decision-making patterns and act proactively, reducing portfolio risk exposure by 23%.

• Engineered and optimized deep learning models including LSTM, Random Forest, and Transformer architectures within Azure Machine Learning pipelines, predicting market-driven stress scenarios and behavioral tendencies with higher precision and improving overall risk insights for investment decisions, increasing predictive accuracy by 28%.

• Orchestrated end-to-end NLP workflows using LangChain and Hugging Face, extracting sentiment, uncertainty, and decision- making cues from portfolio manager notes while integrating GraphQL APIs for seamless access to real-time holdings and liquidity data across financial products.

• Designed and deployed interactive dashboards with Azure Machine Learning and Plotly/Dash, delivering real-time insights into portfolio biases and liquidity forecasts, which improved reporting efficiency and enhanced decision-making timelines.

• Reduced liquidity risk exposure by leveraging predictive analytics to simulate high-pressure redemption scenarios and forecast fund performance, aligning the simulation outputs with business objectives to support proactive financial risk management.

• Deployed optimized machine learning pipelines using Azure ML and Docker while managing model lifecycle with JIRA and collaborating closely with cross-functional teams to ensure seamless integration into production systems and front-end decision tools.

DePaul University, USA Teaching Assistant – Data Science & ML Aug 2024 - Dec 2024

• Facilitated the development of a real-world predictive analytics platform by guiding students in building end-to-end ML pipelines involving data preprocessing, feature engineering, model training, and deployment using Python, Scikit-learn, and Azure Machine Learning, enabling accurate business insights and improving project success rates by 30%.

• Directed collaborative student projects focused on classification and regression modeling by applying Random Forest, XGBoost, and Logistic Regression techniques on real datasets, improving model accuracy and interpretability through optimized hyperparameters and advanced evaluation metrics, boosting overall model accuracy by 25%.

• Enabled effective data visualization and reporting by coaching students on creating interactive dashboards in Power BI and Matplotlib, integrating KPIs, drill-downs, and trend forecasts to support actionable insights and improve reporting efficiency significantly.

• Designed and executed advanced statistical analysis workflows involving hypothesis testing, regression analysis, and A/B testing to strengthen applied data science knowledge, empowering students to interpret results and validate machine learning models effectively.

• Improved student project outcomes by implementing predictive modeling techniques and optimizing deployment strategies, resulting in significant enhancements in decision-making efficiency and strengthening the overall understanding of cloud-based ML applications on Azure.

Hexaware Techno, India Data Science Engineer Dec 2020 – Nov 2022

• Engineered an intelligent user analytics and recommendation platform by leveraging Python, Pandas, and AWS cloud services, enabling personalized content delivery and enhancing overall user engagement by 32% through optimized feature targeting.

• Designed and implemented a recommendation engine using collaborative filtering techniques and predictive modeling approaches to personalize the user journey, significantly improving content discovery and increasing user satisfaction scores by 27%.

• Architected and optimized real-time ETL pipelines using AWS Glue, S3, and Lambda, enabling seamless ingestion, transformation, and processing of large-scale datasets, which enhanced scalability and system performance under heavy traffic.

• Developed and deployed predictive churn models using classification algorithms and advanced data science techniques, empowering the business to proactively identify at-risk users and implement retention strategies effectively.

• Automated data workflows and collaborated with cross-functional engineering teams to integrate machine learning models into production systems, ensuring smooth deployment, robust monitoring, and minimal downtime.

• Enhanced decision-making by developing interactive dashboards and real-time data visualizations in Tableau and Power BI, providing stakeholders with actionable insights into user behavior, platform usage trends, and feature adoption metrics.

• Streamlined cloud-based ML pipelines by orchestrating model training, hyperparameter tuning, and deployment within AWS SageMaker, improving predictive accuracy and aligning analytical outcomes with key business objectives. EDUCATION:

Master’s in Data Science Nov 2024

DePaul University, Chicago, IL, USA

PROJECTS:

Fraudulent Job Posting Detection

• Designed and developed a machine learning solution to classify fraudulent job postings using both textual and meta-data features.

• Engineered and processed large datasets to extract key attributes for fraud detection, ensuring high-quality training data.

• Implemented and compared multiple models including Random Forest, LSTM, XGBoost, and SVM to evaluate performance.

• Achieved 97.3% accuracy with SVM, delivering balanced predictions and significantly improving detection of fraudulent postings.

• Enhanced model robustness through hyperparameter tuning and validation, ensuring scalability for real-world deployment. CERTIFICATIONS:

• Microsoft Certified: Azure Data Scientist

• HackerRank Gold Badge – Python

Contact this candidate