Name : Poornima Vaidya
Role : AI Engineer | Data Engineer
Degree : M.S. Computer Science @ WMU (GPA: 3.7)
Status : Open to Full-time Opportunities
Location : United States- π Building RAG Pipelines and Agentic AI systems
- π Reduced a 2-hour ETL pipeline to 10 minutes at IBM
- π€ Slashed underwriting review time from 10 hrs β 5 mins using GPT-4
- π Published researcher in AI + VR Education (2026)
- π¬ Ask me about LLMs, Data Engineering, MLOps, ETL
- π« Reach me: poornimavaidya2910@gmail.com
- β‘ Fun fact: I mentored 3 engineers at IBM & increased deployment velocity by 30%!
- Engineered high-scale ETL pipelines managing 100K+ records across Oracle EBS suite for a 51,000+ employee global client
- Containerized legacy systems using Docker & Python; reduced execution latency by 98% (2 hrs β 10 mins)
- Automated CI/CD pipelines using Git & Shell, boosting deployment velocity by 70%
- Built anomaly detection system in Python reducing data errors by 60%
- Mentored 3 associate engineers; recognized by IBM leadership for outstanding contribution
- Architected a RAG pipeline using LangChain, GPT-4 & PostgreSQL β slashed underwriting review from 10 hrs β 5 mins
- Built a full-stack AI chatbot in React & Python using NLP and NER to autonomously handle loan inquiries
- Developed an "Agent Builder" app enabling 30+ stakeholders to deploy custom AI tools from project charters
- Improved Power BI reporting accuracy by 40% across 100+ evaluations using automated data pipelines
- Co-authored published paper: Gamified Immersive Learning Experiences for Promoting AI Readiness across STEM Disciplines (2026)
- Integrated real-time ML model behavior into Unity & C# VR systems, improving student engagement by 30%
- Built ML models in Python (TensorFlow, PyTorch) for classification, predictive analytics and EDA
π©βπ» Languages
βοΈ Cloud & Data Platforms
βοΈ Data Engineering
π€ AI / ML
ποΈ Databases & Visualization
| π Deepfake Detection | π Data Warehousing |
|---|---|
| Xception + LSTM model achieving 95% accuracy | Medallion Architecture, 60% less manual effort |
| Face extraction, augmentation, real vs manipulated classification | Star schemas, automated ETL, 8+ Power BI dashboards |
| Python Β· TensorFlow Β· OpenCV Β· CNN Β· LSTM | SQL Server Β· SSIS Β· Power BI Β· Star Schema |
Collaborated with a large team under NASA's open science initiative
- π‘ Analyzed the 2024 Maui floods using NASA GPM IMERG and MODIS satellite data
- πΊοΈ Delivered flood-risk maps and animations for vulnerable communities and city planning
- π Stack:
PythonΒ·pandasΒ·xarrayΒ·geopandasΒ·rasterioΒ·matplotlibΒ·OpenStreetMapΒ·U.S. Census Data
| Degree | Institution | Year | Score |
|---|---|---|---|
| M.S. Computer Science | Western Michigan University, Kalamazoo MI | May 2026 | GPA: 3.7/4.0 |
| B.Tech Electronics & Telecommunications | Dr. Babasaheb Ambedkar Marathwada University | May 2019 | GPA: 3.4/4.0 |
Relevant Coursework: Generative AI Β· Advanced Artificial Intelligence Β· Machine Learning Β· Pattern Recognition Β· Big Data Β· DBMS
- β Microsoft Certified: Azure Fundamentals (AZ-900)
