Skip to content
View Sahithibrunda8's full-sized avatar

Block or report Sahithibrunda8

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.

Content in all repositories owned by your account will be closed.
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
Sahithibrunda8/README.md

Hi, I'm Sahithi Brunda 👋

Data Scientist building end-to-end machine learning systems from data pipeline to deployed API, not just model training.

About Me

I build and deploy ML systems that go beyond a notebook real APIs, real statistical rigor, real verification of my own results. My recent work includes cost-sensitive fraud detection, statistical A/B testing frameworks, credit risk modeling, and hybrid retrieval-augmented generation (RAG) systems.

Technical Skills

Python | SQL | Machine Learning | Statistical Modeling | XGBoost | Scikit-learn | SHAP | Hypothesis Testing | A/B Testing FastAPI | Docker | LangChain | ChromaDB | BM25 | RAG | Streamlit | Git | GitHub Actions | pytest | Pandas | NumPy

Featured Projects

🔍 Fraud Detection System Cost-sensitive XGBoost model (0.867 PR-AUC) deployed live with FastAPI + Docker, SHAP explainability, drift monitoring. Live API →

📊 A/B Testing & Experimentation Framework 7-stage statistical pipeline — SRM checks, power analysis, Bonferroni-corrected guardrail metrics — on real messy data cleaned via SQL.

💳 Credit Risk Scoring Leakage-safe credit default prediction with SQL-based EDA and stratified cross-validation.

🔎 Hybrid RAG Document Intelligence BM25 + dense vector retrieval fused with Reciprocal Rank Fusion, citation-grounded generation.

📫 Reach me: LinkedIn · sahithibrunda@gmail.com

Pinned Loading

  1. A-B-testing-tool A-B-testing-tool Public

    A/B testing evaluation framework with SQL-based data cleaning, SRM/power checks, Bonferroni-corrected guardrail metrics, and novelty-effect detection built on deliberately messy raw data.

    Python 1

  2. Credit-Risk-Scoring Credit-Risk-Scoring Public

    End-to-end credit default prediction on Lending Club loan data SQL-based EDA, leakage-safe stratified cross-validation, and model selection for interpretability

    Python 1

  3. fraudsim-risk-engine fraudsim-risk-engine Public

    Real-time fraud detection system with a self-built transaction simulator, cost-sensitive XGBoost model, SHAP explainability, and drift monitoring deployed live with FastAPI + Docker on Render.

    Python 1

  4. Hybrid-RAG-System Hybrid-RAG-System Public

    Hybrid RAG system (BM25 + Qdrant dense search + RRF) with cross-encoder reranking, ablation-tested retrieval, page-cited document upload, and full CI/CD.

    Python 1