Skip to content
View Hrushitha12's full-sized avatar
  • Charlotte, NC

Block or report Hrushitha12

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
Hrushitha12/README.md

Hi, I'm Hrushitha 👋

Data Scientist & AI Researcher with experience building production ML systems, LLM evaluation frameworks, and agentic AI pipelines. I turn complex data problems into deployed, measurable solutions.

MS Computer Science (Data Science) from UNC Charlotte · Previously Data Scientist at CVS Health and TCS · AWS Certified Data Engineer · 2 IEEE publications

Portfolio LinkedIn Email


🔬 What I Work On

  • LLM Evaluation & Robustness — Built RobustnessPilot, an automated framework that executed 38 evaluation runs across 3 LLMs (14B–70B params), generating 663 tests across 23 failure modes. First-authored paper submitted to IEEE SRDS 2026.
  • Agentic AI Systems — Architected a 5-agent API test suite generator using CrewAI, MCP, and A2A Protocol with GPT-4o/Claude Sonnet routing via LiteLLM. Exposed as an MCP server for Claude Desktop tool invocation.
  • Production RAG & NLP — Built a legal research assistant over 22,809 court opinion vectors using LangChain, LangGraph, and Qdrant — 277% retrieval improvement, ~30% reduction in hallucinated citations.

📄 Publications

Paper Venue Year
LLM-Based Robustness Testing of Microservice Applications (First Author) IEEE SRDS 2026 2026
Crop Yield Prediction using ML & Deep Learning [IEEE] (https://ieeexplore.ieee.org/document/10689872) 2024
Legal Precedent Retrieval using Sentence Embeddings and Clustering arXiv:2406.01609 2024

🛠️ Tech Stack

ML & AI: Python · PyTorch · TensorFlow · Scikit-Learn · XGBoost · PySpark MLlib · Hugging Face Transformers

LLMs & Agents: LangChain · LangGraph · CrewAI · MCP · A2A Protocol · RAG · Prompt Engineering · LLM Evaluation · RAGAS

Data: SQL · Pandas · NumPy · Snowflake · Spark SQL · Tableau · Streamlit · Matplotlib

Infrastructure: FastAPI · Docker · AWS (SageMaker, Lambda, S3) · MLflow · Qdrant · Git · CI/CD · Airflow


🏆 Certifications

AWS Certified Data Engineer


💻 Tech Stack:

Java JavaScript Python R TypeScript C C++ Markdown Azure Google Cloud Oracle Apache Spark Apache Hadoop NodeJS OpenCV React Streamlit Jenkins Apache Apache Maven MySQL MongoDB Keras Matplotlib mlflow NumPy Pandas Plotly PyTorch scikit-learn Scipy TensorFlow GitHub Actions Git GitHub Docker Power Bi

Pinned Loading

  1. Portfolio Portfolio Public

    Personal portfolio website — Data Scientist & AI Researcher | React, Vite, Tailwind CSS | GitHub Pages

    Python

  2. agentic-test-generator agentic-test-generator Public

    5-agent AI system using CrewAI, MCP, and A2A Protocol to autonomously generate API test suites from OpenAPI specs — GPT-4o/Claude Sonnet via LiteLLM with LangSmith tracing

    Python

  3. legal-rag-api legal-rag-api Public

    Production RAG API for legal precedent retrieval — extends arXiv:2406.01609

    Python

  4. llm_robustness_test_diversity llm_robustness_test_diversity Public

    RobustnessPilot — Automated LLM evaluation framework testing 3 models (14B–70B params) across 23 failure modes with 7 prompt strategies | First-author paper submitted to IEEE SRDS 2026

    Java

  5. Machine-Learning-Model-Reliability-Monitoring-Dashboard Machine-Learning-Model-Reliability-Monitoring-Dashboard Public

    Production ML monitoring dashboard tracking performance drift, data drift, and prediction drift across Classification, Regression, NLP, and Computer Vision — SHAP, LIME, KS tests

    Jupyter Notebook

  6. smart-parking-analytics-dashboard smart-parking-analytics-dashboard Public

    End-to-end ML pipeline with PySpark MLlib and Scikit-Learn (0.71 F1, 0.79 AUC) + interactive Streamlit dashboard for parking analytics on 100K+ records

    Jupyter Notebook