Open to opportunities

Sabari Murugan S

@sabarimurugans

Message

Machine Learning Engineer focused on reproducible, deployment-ready ML models and scalable insights.

India

Message

What I'm looking for

I’m looking for a role where I can build reproducible, deployment-ready ML pipelines, work with real data at scale, and ship models that stakeholders can use—while improving my skills in end-to-end modeling, evaluation, and communication.

I’m a Machine Learning Engineer focused on reproducible workflows, deployment-ready models, and scalable data solutions. I build and evaluate ML/DL models, design robust preprocessing pipelines, and turn complex datasets into actionable insights.

In my internships and projects, I developed an Influenza RNA Similarity model using 3-mer (k-mer) features and cosine similarity, including DNA→RNA processing and segment-wise matching, and I deployed an interactive Gradio app on Hugging Face Spaces for real-time analysis. I also built a Heart Stroke Prediction model reaching 92% accuracy with thorough preprocessing and model evaluation, and I created a deployment-ready Concrete Strength Prediction pipeline using ensemble regressors (Random Forest, SVR, Bagging, Voting Regressor) with 0.92 R²—while communicating results through dashboards in Power BI.

Experience

Work history, roles, and key accomplishments

Machine Learning Intern

Anjana Infotech

Nov 2025 - Apr 2026 (5 months)

Built an Influenza RNA similarity model using 3-mer features and cosine similarity, including DNA→RNA preprocessing, k-mer feature scaling, and segment-wise matching to return top similar strains. Deployed an interactive Gradio app on Hugging Face Spaces for real-time analysis.

Python Pandas NumPy scikit learn NLTK Biopython Feature Engineering Cosine Similarity Gradio

Data Analytics Intern

Anjana Infotech

May 2025 - Present (1 year 1 month)

Developed a heart stroke prediction model using supervised ML, achieving 92% accuracy through data cleaning, preprocessing, missing-value handling, outlier detection, and normalization/encoding. Performed EDA and model evaluation using confusion matrix, ROC/precision-recall curves, and feature importance, then documented results for mentors.