Saksham Tyagi

Data Scientist @JPMorganChase

TX, US

Developed and deployed Python machine learning pipelines using Scikit-learn, XGBoost, and TensorFlow, improving credit risk prediction accuracy by 28% across retail and corporate banking portfolios while ensuring compliance with regulatory standards. Implemented LangChain and Retrieval-Augmented Generation (RAG) frameworks with GPT-4 and Claude to automate financial document summarization and anomaly detection, reducing manual processing time by 40% and increasing reporting accuracy. Performed Natural Language Processing (NLP) tasks, including Named Entity Recognition (NER) and text classification, on unstructured financial documents and client communications, generating datasets that improved model performance metrics by 22%. Engineered real-time streaming analytics solutions using AWS Kinesis, Redshift, and Google Cloud Platform (GCP) services, enabling live transaction monitoring and alerting, which prevented potential fraud losses of over $3M within the first six months of implementation. Orchestrated end-to-end MLOps pipelines with CI/CD for multiple predictive models, ensuring model reproducibility, automated testing, and rapid deployment, which decreased model deployment time from weeks to 48 hours. Designed and visualized actionable dashboards in Tableau for portfolio managers and regulatory teams, integrating LLM-based insights and financial KPIs, which improved decision-making efficiency by 35% for high-value investment portfolios.

Meerut Institute of Engineering and Technology(MIET)

Bachelor of Technology - BTech

The University of Texas at Arlington

Master's degree