Data Scientist & ML Engineer
I build production-focused machine learning systems with strengths in modeling, MLOps, LLM/RAG pipelines, and data engineering. My work emphasizes clean architecture, reliable data pipelines, and measurable impact.
- Graduate Research Assistant at WVU leading energy analytics, automation, and ML-driven decision support tools.
- Strong background in applied machine learning, LLM-based retrieval systems, and Python software engineering.
- Skilled at turning ambiguous business problems into scalable, maintainable solutions.
Languages: Python, SQL, TypeScript, JavaScript
ML/AI: XGBoost, CatBoost, Random Forest, Logistic Regression, SVM, Optuna, SMOTE, SHAP
LLM/RAG: FAISS, pgvector, Docling, embedding pipelines, retrieval optimization
MLOps: Docker, FastAPI, MLflow, GitHub Actions, Streamlit, Poetry, uv
Data Engineering: ETL processes, PostgreSQL, Elastic Stack, data modeling
Frontend (for tools): React, shadcn/ui, TailwindCSS, DataTables.net
Local, privacy-preserving RAG pipeline for energy-audit PDFs using Docling, FAISS, pgvector, FastAPI, and Streamlit.
Modular ingestion → embedding → indexing → retrieval workflow using FAISS and local LLMs.
Python + Elastic pipeline classifying log events into NIST 800-53 controls, with FastAPI backend and a lightweight React dashboard.
End-to-end ML workflow: preprocessing, feature engineering, SMOTE, Optuna tuning, SHAP explainability, and reporting.
- Machine Learning & MLOps
- LLM/RAG systems and AI agents
- Data engineering for analytics and automation
- Applied ML in energy, manufacturing, and enterprise AI
- LinkedIn: linkedin.com/in/aksh-ay06
- Email: [email protected]
- Portfolio: akshayxcode.dev
“You don't need to be a genius to change the world. Just be consistently curious.”