B
Bala Krishna Reddy Ragannagari
About
Detail
Remote, Oregon, United States
Résumé
Jobs
verified_user
0% verified
AI/ML EngineerSocureAug 2024 - Current (2 years 3 months)• Spearhead development of deep learning models using Transformers, CNNs, and LSTMs for fraud detection, improving risk
prediction accuracy by 30% across real-time identity systems processing 3M+ transactions daily.
• Architect multimodal ML pipelines leveraging Python, PyTorch, and TensorFlow across tabular, NLP, and image data,
optimizing feature engineering and inference, reducing latency and increasing model performance.
• Own end-to-end machine learning lifecycle, including feature engineering, training, evaluation, deployment, and model
monitoring, scaling production systems to support 2M+ daily transactions, improving precision-recall performance.
• Drive real-time inference systems using Ray, LangChain,- Data ScientistTata Consultancy ServicesJan 2021 - Jul 2023 (2 years 7 months)• Analyzed 1M chatbot interactions using Python, SQL, and Apache Spark on AWS EMR, identifying misclassification patterns,
collaborating with product and engineering teams to optimize intent routing, increasing automation rate by 20%.
• Built and optimized BERT-based intent classification models on large conversational datasets, deploying training pipelines on
AWS SageMaker, tuning precision-recall thresholds with stakeholders to improve F1-score.
• Engineered confidence-based decision pipelines using model probability outputs in Python, enabling dynamic response
routing, partnering with product managers to reduce fallback rates and improve chatbot reliability.
• Conducted exploratory data analysis and error di - Junior Data ScientistRazorpayFeb 2020 - Dec 2020 (11 months)• Extracted, processed, and analyzed 1M+ transaction records using SQL, Python, Pandas, and NumPy, performing joins,
aggregations, and advanced preprocessing to improve data quality, accessibility, and reliability for fraud detection workflows.
• Developed machine learning models using Scikit-learn (Logistic Regression, Random Forest, Isolation Forest), enabling
early fraud detection and contributing to $120K annual fraud loss prevention through improved model performance.
Education
verified_user
0% verified
- Master of Science in Data ScienceThe George Washington UniversityAug 2023 - May 2025 (1 year 10 months)Washington, DC
Projects (professional or personal)
verified_user
0% verified
- CCredit Card Fraud Detection• Developed a fraud detection pipeline on 1M+ transactions using Airflow and AWS S3
• Trained an XGBoost model achieving 82% recall for fraud detection
• Implemented MLflow and DVC for experiment tracking and data versioning - MMedical Conversational AI Chatbot• Built a medical RAG system using clinical documents for accurate information retrieval
• Developed hybrid retrieval (vector search + BM25) with LLM-based response generation
• Tracked experiments with MLflow and optimized performance using caching and CI/CD