Satya Dineswara Reddy Satti

About Satya Dineswara Reddy Satti

Machine Learning Engineer with around 4 years of specialized experience developing and…

Experience

1. Software Engineer Ai

   Northern Trust

   Jan 2025 — Present · US

   Designed and fine-tuned a domain-specific Large Language Model (LLM), using Hugging Face Transformers and PyTorch, for summarizing quarterly trust and asset management reports, automating the extraction of 85% of key financial risk metrics-Architected and managed a production MLOps pipeline on AWS (SageMaker Pipelines and Model Monitor), accelerating the continuous integration and delivery (CI/CD) of updated models by 40% while ensuring 99.9% endpoint uptime-Developed a high-frequency Time Series Forecasting model utilizing LSTM Neural Networks (TensorFlow/Keras) to predict potential liquidity shortfalls, reducing false positive risk alerts by 30% compared to traditional ARIMA models-Containerized ML inference services (XGBoost fraud models) using Docker and deployed them on a secure Kubernetes cluster via FastAPI, achieving an average prediction latency of <50 milliseconds under peak load-Implemented SHAP and LIME frameworks to provide regulatory-compliant explainability for a high-stakes credit risk classification model, directly influencing the approval process for over $500M in commercial loans-Engineered feature pipelines using Apache Spark (PySpark) to process 10TB+ of raw transaction and market data from Snowflake, reducing feature calculation latency for real-time model serving by 55%-Conducted multi-variant A/B Tests on deployed models via AWS SageMaker, iteratively increasing the AUC-ROC score for transaction anomaly detection from 0.88 to 0.94 over two sprints.

Education

• Lovely Professional University

  Bachelor of Technology - BTech

  2018 — 2022

• Illinois Institute of Technology

  Master of Science - MS

  2023 — 2025