Summary
Overview
Work History
Education
Skills
Websites
Certification
Publications
Timeline
Generic

Ashish Verma

Union City,CA

Summary

Seasoned Applied Scientist and Data Engineering Leader with 14+ years of experience at PayPal and enterprise organizations, specializing in designing and scaling data pipelines, AI/ML systems, and deploying LLM-driven solutions. Proven track record in fraud detection, anomaly detection, and RAG implementations using AI Agents.

Overview

15
15
years of professional experience
1
1
Certification

Work History

Applied Scientist & Lead Data Engineer

PayPal Inc
02.2022 - Current
  • Agent based data ingestion pipeline for automate schema evolution for M&A projects.
  • Multi agent based based workflow for daily batch lag detections.
  • MCP based BigQuery dataset summarizer.
  • Engineered anomaly detection system using VAE-BiLSTM and CNN-BiLSTM models on Kafka, Spark, and GCP, reducing detection time by 40% and enhancing system security.
  • Implemented NLP-driven RAG system with LangChain and LlamaIndex docs.reltio.com as external MDM system for UEC program.
  • LLM Fine tuning using LoRA and QLoRA techniques.
  • Developed interactive dashboards in Tableau and Power BI to monitor key operational metrics and ML model performance.
  • Architecture and led development of scalable batch and real-time data pipelines using Apache Spark and Kafka, improving data latency and ensuring high data quality.
  • Mentored and guided cross-functional teams on ML best practices, feature engineering, and production deployments.

Data Engineer

Kforce Inc
02.2020 - 02.2022
  • Built and optimized real-time data ingestion pipelines with Kafka Streams and GCP Dataflow for AI-driven applications.
  • Collaborated on time-series forecasting models using TensorFlow, Spark ML, and PyTorch, improving forecast accuracy by 30%.
  • Automated data workflows and containerized microservices with Docker and Kubernetes, reducing deployment time by 30%.
  • Established a centralized feature store using PySpark and BigQuery, streamlining the ML model training and deployment lifecycle.

Data Engineer

Cognizant Technology
03.2014 - 02.2020
  • Designed ETL workflows with Python Frameworks and re-architected Spark-based pipelines to migrate large-scale financial datasets to cloud environments, improving processing efficiency by 20%.
  • Deployed classification models for fraud detection using Scikit-learn and statistical analysis, enhancing detection capabilities in financial systems.

Software Engineer

Wipro Technologies
11.2010 - 03.2014
  • Developed enterprise data warehouse solutions and automated reporting processes using PL/SQL and Python, reducing manual reporting by 50%.

Education

PhD - Machine Learning

University of York
London
12.2031

Master of Science - Data Science & Analytics

Old Dominion University
Norfolk, VA
03.2025

Bachelor of Technology - Computer Science & Engineering

Kanpur University
India
06.2010

Skills

  • Proficient in Agentic AI tools
  • Tools & Frameworks: TensorFlow, PyTorch, Scikit-learn, Spark ML, LangChain, LlamaIndex
  • Machine Learning & AI: Regression, Classification, Clustering, Deep Learning (CNN, RNN), NLP, LLMs, RAG, Anomaly Detection
  • Data Engineering: Apache Spark, Kafka, Databricks, Airflow, BigQuery, Pub/Sub, Hadoop
  • Cloud & DevOps: GCP (BigQuery, Dataflow, Pub/Sub, SageMaker), AWS, Kubernetes, Docker, CI/CD (Jenkins, Git)
  • Programming: Python, SQL, R, C, CUDA
  • Visualization: Tableau, Power BI, Plotly, Matplotlib, Seaborn

Certification

  • Google Cloud Certified Professional Data Engineer
  • Google Cloud Certified Professional Machine Learning Engineer
  • Azure Certified Data Scientist
  • Data bricks Spark Certified Professional
  • Patent: AI Fraud Investigation Assistant (PayPal) in process.
  • Patent: Cross-Border Payment Navigator (PayPal) in process.

Publications

  • Energy-Aware Prompt Optimization for Large Language Models: Balancing Accuracy, Cost, and Sustainability. TechRxiv. https://www.techrxiv.org/users/954699/articles/1323739
  • Variational Autoencoder and BiLSTM-based Anomaly Detection for Beam Stability in Spallation Neutron Sources. TechRxiv. https://www.techrxiv.org/users/954699/articles/1323733

Timeline

Applied Scientist & Lead Data Engineer

PayPal Inc
02.2022 - Current

Data Engineer

Kforce Inc
02.2020 - 02.2022

Data Engineer

Cognizant Technology
03.2014 - 02.2020

Software Engineer

Wipro Technologies
11.2010 - 03.2014

Master of Science - Data Science & Analytics

Old Dominion University

PhD - Machine Learning

University of York

Bachelor of Technology - Computer Science & Engineering

Kanpur University
Ashish Verma