Skip to content
View Aditya20029's full-sized avatar
2 followers · 0 following

Block or report Aditya20029

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Aditya20029/README.md

👋 About Me

I'm an AI Engineer based in Fairfax, VA, finishing my M.S. in Data Analytics at George Mason University (May 2026). I build production RAG systems, LLM orchestration pipelines, and hybrid retrieval engines that ship to real users, not notebooks.

  • 🛰️ Currently leading backend on Arctic Policy Assistant, a legal intelligence RAG platform for JAG analysts. Built at the GMU DAPSE 3.0 capstone with NSI (National Security Innovations).
  • 🎓 Graduate Teaching Assistant for AIT-580 Data Analytics at George Mason, supporting 50+ graduate students.
  • 🏅 5x AWS Certified including Generative AI Developer Professional and ML Engineer Associate.
  • 🛠️ Comfortable across the stack: Python, FastAPI, Docker, React, Next.js, AWS, Postgres, Snowflake.
  • ✉️ Reach me at reachaditya29@gmail.com.

🚀 Featured Work

🛰️ Arctic Policy Assistant

Production RAG platform that lets JAG analysts query 1,630 policy sources across 21 countries in seconds instead of hours.

  • Hybrid search engine (SQLite FTS5 + FAISS + Reciprocal Rank Fusion) with authority-weighted reranking
  • nDCG@5 = 0.832, Precision@5 = 0.954 on curated eval suite
  • 7-stage async LLM orchestration on the OpenAI SDK with checkpoint recovery, per-model circuit breakers, Langfuse observability
  • ~80% API cost reduction by routing simple queries to GPT-5-nano
  • Shipped on GMU OpenStack: FastAPI + SSE streaming, Docker behind nginx, token auth, rate limiting
  • Validated by 1,481 automated tests, certified for NSI hand-off

Python FastAPI OpenAI FAISS Docker Langfuse nginx

🔥 Wildfire Risk Prediction

Geospatial ML pipeline fusing MODIS satellite fire data, NOAA climate variables, and NDVI vegetation indices.

  • AUC-ROC = 0.99 with XGBoost
  • GeoPandas risk-zone maps for proactive response

Python XGBoost GeoPandas MODIS NOAA

📊 SLA Breach Prediction (ITSM Analytics)

End to end ML pipeline over 5,000 synthetic tickets. XGBoost beat baselines for resolution-time and breach prediction. Power BI dashboard with drill-down by priority, assignee, and SLA status.

Python XGBoost Scikit-learn Power BI

🌐 Obesity Risk Analytics

Cloud-native pipeline (S3 to Glue to RDS) over CDC BRFSS data. Regression, Random Forest, and ARIMA models to surface county-level obesity trends and socioeconomic risk factors.

AWS S3 Glue DataBrew RDS R Python

🚗 Real-Time License Plate Detection

YOLO localization combined with Tesseract OCR for near-real-time plate extraction across varied formats.

YOLO Tesseract OpenCV Python

🧰 Tech Stack

Languages

Python SQL TypeScript JavaScript R Bash

AI / ML and Data Science

OpenAI Anthropic LangChain Langfuse Hugging Face FAISS

TensorFlow PyTorch scikit-learn XGBoost OpenCV YOLO

Pandas NumPy PySpark Jupyter

Backend, Cloud, and Infra

FastAPI Flask Node.js Docker nginx Prometheus GitHub Actions

AWS Azure GCP Snowflake Databricks

Databases and Frontend

PostgreSQL MySQL SQLite MongoDB Power BI

React Next.js Tailwind HTML5 CSS3

Tools

Git GitHub VS Code Linux Figma Postman Vercel

🏅 AWS Certifications (5x)

Verify any certification at aws.amazon.com/verification

📈 GitHub Stats

profile stats streak stats

most used languages by commits repos per language

productive time

trophies

activity graph

🎓 Education
George Mason University M.S. Data Analytics, 2024 to 2026
St. Peters Engineering College B.Tech CSE (AI and ML), 2020 to 2024

🤝 Let's Connect

linkedin gmail github phone

📍 Fairfax, VA • 🟢 Open to AI / ML Engineer roles • 🛂 F1 STEM OPT Eligible

"Build things that ship. Measure what matters. Keep learning."

footer

Pinned Loading

  1. Arctic-Policy-Assistant Arctic-Policy-Assistant Public

    This is an Arctic Policy Intelligence Engine — a web-based tool that helps analysts understand and evaluate geopolitical policies related to the Arctic region.

    Python

  2. pulse pulse Public

    Pulse: real-time 3D global news intelligence globe with Claude Opus 4.7 briefings, GDELT/Reddit/RSS feeds, and a mission-control HUD.

    TypeScript 1

  3. Ticket-Resolution-Time-Prediction-ITSM- Ticket-Resolution-Time-Prediction-ITSM- Public

    End-to-end ITSM analytics project predicting ticket resolution time and SLA breaches using synthetic data. Includes data generation, EDA, regression and classification models (Linear, XGBoost), eva…

    Jupyter Notebook

  4. authflow authflow Public

    Prior authorization automation modeling the HL7 FHIR Da Vinci CRD, DTR, and PAS workflow. Next.js 14, TypeScript, Prisma, and the Claude API. Synthetic-data demo.

    TypeScript

  5. enhancing-Visual-Clarity-and-Streamlining-Information-Lookup-Number-Plates enhancing-Visual-Clarity-and-Streamlining-Information-Lookup-Number-Plates Public

    License plate detection using Optical Character Recognition (OCR), Python, and YOLO (You Only Look Once) involves leveraging YOLO's object detection capabilities to locate license plates within ima…

    Jupyter Notebook

  6. Obesity-and-Its-Drivers-A-BRFSS-Analysis Obesity-and-Its-Drivers-A-BRFSS-Analysis Public

    Analyzes BRFSS data to explore links between physical activity, diet, socioeconomic factors, and obesity in the U.S. Utilizes AWS (S3, Glue DataBrew, RDS), Python, R, and SQL for data cleaning, ana…

    Python