R#hît @nànt#àn
D@ta $c!ent!st & A! Eng!neer
Model trained on 4+ years of real-world data — converging on AI/ML, product analytics & GenAI problems with GCP, AWS, LLMs & real-time inference pipelines. val_loss → 0.0000 ✓
About Me
Data Scientist with 4+ years building production ML systems across e-commerce, non-profit, and enterprise domains. Specializing in LLM-powered applications, product analytics, real-time anomaly detection, and scalable data engineering pipelines using Python, PySpark, GCP, AWS, and Snowflake.
With a background spanning Mechanical Engineering (B.Tech from IIITDM Kancheepuram) to Information Systems (MS from University of Maryland — Smith School of Business), I bring a rare blend of engineering rigor and business context to every data problem.
Experience
Data Scientist Consultant
CurrentInvision Global Tech Inc
Full-time · United States
Data Scientist
Community Dreams Foundation
Full-time · Remote
- ▸Built an AI-powered Legal & Compliance Assistant using GPT-4o with a RAG pipeline on LangChain, Pinecone, and ANN search — cutting manual review time by 40%
- ▸Deployed end-to-end ML pipelines on GCP Vertex AI and Dataproc with automated hyperparameter tuning via MLflow — improving demand forecasting accuracy by 15%
- ▸Implemented MLOps workflows using Vertex AI, Cloud Build, and GitHub Actions for automated model versioning, drift detection, and CI/CD across environments
- ▸Built a real-time fraud and anomaly detection system using Pub/Sub, Dataflow (Apache Beam), and XGBoost — reducing undetected fraud by 20%
- ▸Engineered scalable data pipelines with Dataflow, Composer (Airflow), and BigQuery — cutting ETL latency by 60%
- ▸Developed LLM-based apps for document summarization and Q&A using OpenAI APIs and Vertex Matching Engine for semantic search
- ▸Created predictive donor churn models with TensorFlow and Scikit-learn, deployed on Vertex AI with Looker Studio dashboards
Financial Analyst
The Premiere Group
Full-time · Columbia, MO · On-site
Technical Consultant – Course Renewal Automation
University of Maryland – Extended Studies
Internship · College Park, MD · Remote
Graduate Assistant
University of Maryland
Part-time · College Park, MD · On-site
Data Scientist
Kameleon Technologies
Full-time · Chennai, India
- ▸Engineered a real-time fraud detection pipeline using Neo4j graph database and XGBoost — processing 5M+ daily transactions with an 18% reduction in false positives
- ▸Built scalable ETL/ELT pipelines on PySpark and AWS (S3, Glue, EMR) to process terabytes of financial data — reducing pipeline processing time by 35%
- ▸Developed customer segmentation and churn prediction models using ensemble methods — driving a 12–15% improvement in customer retention across key segments
- ▸Established MLOps practices with MLflow experiment tracking, SageMaker model registry, and automated retraining workflows for production model governance
- ▸Delivered executive-facing Tableau dashboards for transaction monitoring, KPI tracking, and fraud trend analysis — adopted across operations and risk teams
Skills
Languages & Libraries
ML & AI
Data Engineering
Cloud & Infra
BI & Visualization
Databases & Search
Projects
Gym Aesthetic Trap
NLP research project using LDA topic modeling to analyze online discourse around SARMs and steroid usage — uncovering themes, risk perception patterns, and community sentiment from bodybuilding forums.
Thermal Error ML Modeling
Published Springer research on machine learning compensation strategies for thermal deformation in precision machine tools — achieving state-of-the-art accuracy in error prediction.
Education
Master of Science – Information Systems
University of Maryland
Robert H. Smith School of Business
📍 College Park, MD
Bachelor of Technology – Mechanical Engineering (Smart Manufacturing)
IIITDM Kancheepuram
Indian Institute of Information Technology Design & Manufacturing
📍 Kancheepuram, India
Certifications
Neo4j Graph Data Science Certification
Neo4j
AWS Certified AI Practitioner (AIF-C01)
Amazon Web Services
BCG Data Science Job Simulation
Boston Consulting Group × Forage
Publications
Mathematical Modeling of Thermal Error Using Machine Learning
Springer · Oct 6, 2022Research on thermal error modeling in machine tools using machine learning algorithms to identify the most effective compensation strategies for linear expansion and deformation caused by heat inputs from internal and external sources.
Contact
Ready to deploy?
Let's ship something intelligent.
Training complete — now seeking inference in the real world. Open to Data Scientist, AI Engineer, and ML Engineer roles. Whether you have a full-time opportunity or just want to talk about model architectures — my context window is open.