Soumya Pednekar
4+ years of experience Data Engineer | LLM & RAG Systems | Credit Risk, Fraud Detection, Anomaly Detection | AWS, Spark, ML, MLOps | AWS Certified | Google Data Analytics Certified | MS in IS @UMBC
- Role
- Data Engineer at Morgan Stanley
- Location
- Baltimore, MD, US
- LinkedIn followers
- 500 followers
About Soumya Pednekar
Data Engineer with 4+ years of experience building scalable data platforms, ML pipelines, and LLM-powered systems within financial services and enterprise consulting environments. Proven expertise in Python, SQL, cloud-native architectures (AWS, Azure, GCP), big data ecosystems (Spark, Hadoop), and end-to-end MLOps, including CI/CD, Docker, and production deployment. Specialized in AI/LLM engineering, RAG pipelines, predictive modeling, fraud detection, and real-time anomaly detection to drive risk analytics, compliance automation, and business intelligence. Demonstrated success delivering highimpact solutions processing millions of records daily while improving forecast accuracy, operational efficiency, and data quality. AWS Certified AI Practitioner with strong experience translating complex model outputs into actionable insights for technical and non-technical stakeholders.
Experience
Data Engineer
Mar 2025 — Present · US
Built LLM-powered RAG (Retrieval-Augmented Generation) pipelines for credit policy summarization, risk narrative generation, compliance alerting, reducing manual review workloads by 35% and integrating human-in-the-loop validation.• Designed AI-driven automation agents for report generation and document triage leveraging GPT and Gemini models, increasing throughput and efficiency for risk and credit teams.• Developed predictive credit-scoring and delinquency models using Logistic Regression and XGBoost, improving forecast accuracy by 18% and strengthening portfolio risk insights.• Implemented a fraud detection pipeline using anomaly detection and statistical modeling, lowering false positives by 22% and enhancing real-time transaction monitoring.• Orchestrated end-to-end AWS ML and data infrastructure, including Glue, Redshift, Lambda, and S3, processing 2M+ daily records and reducing data refresh latency by 40%.• Delivered model explainability and interpretability using SHAP, SQL, and Matplotlib, enabling risk and compliance teams to make transparent, data-driven credit and operational decisions.
Education
New Horizon Scholars School - India
High school, Science
University of Mumbai
Bachelor of Science - BS, Information Technology
2020 — 2023
University of Maryland Baltimore County
Master's degree, Information Systems
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.