KRUSHNA YEOLE

AI & ML Engineer | Data Scientist | Python Developer

+91-9518347964 yeolekrushnar@gmail.com LinkedIn GitHub Portfolio

Professional Summary

Computer Engineering student with hands-on experience in Python, AI/ML, Data Science. Skilled in building predictive models, performing exploratory data analysis, and deploying ML solutions. Passionate about AI/ML, NLP, and leveraging data-driven insights for business decision-making. Strong problem-solving abilities, open-source contributions, and experience in deploying end-to-end applications.

Academic Qualifications

B.E. Computer Engineering

Zeal College of Engineering and Research, Pune

2022–2026 GPA: 7.22/10

HSC (Science)

J.H.S.H College, Nagar

2020–2021 89.00%

SSC (Science)

B.J.S School, Wagholi

2018–2019 84.40%

Technical Skills

Programming & Scripting

Python C++ (Basics)

Data Science & Analytics

Pandas NumPy Matplotlib Seaborn

Machine Learning

Logistic Regression Random Forest Decision Tree SVM KNN

Deep Learning & NLP

Transformers LLMs Natural Language Processing

Model Development

Feature Engineering Model Training & Optimization Evaluation Metrics

Tools & Frameworks

HuggingFace Scikit-learn FastAPI Django

Databases

MySQL SQLAlchemy

Deployment

Web Deployment HTML, CSS, JS

Featured Projects

Language Analyzer: Marathi – Grammar Correction System

View Project
NLP Transformers FastAPI
  • Built a Marathi grammar correction model using IndicBART (ai4bharat) fine-tuned on 200k+ parallel incorrect–correct sentence pairs
  • Data preprocessing: text cleaning with regex filters for non-Devanagari, extra spaces, noise removal
  • Model training: Seq2Seq using HuggingFace Transformers with train/validation split (180k/20k), optimized via AdamW
  • Deployed end-to-end system with FastAPI backend, SQLAlchemy, and responsive frontend
  • Provides grammatical correction for Marathi text in under 1-5 seconds using beam search decoding

Tasty Bytes – Recipe Popularity Prediction

View Project
Machine Learning Predictive Analytics
  • Implemented ML models (Logistic Regression, Random Forest, SVM) to predict recipe popularity from nutritional data
  • Achieved 76% accuracy, 80% precision with Logistic Regression
  • Delivered insights into high-traffic recipe categories, improving website traffic strategies

Certifications

Oracle Cloud Infrastructure 2025 Certified Data Science Professional

Oracle

Validates advanced skills in EDA, feature engineering, model training, optimization, deployment, and ML lifecycle management in cloud.

Personal

Strengths

Problem-Solving Teamwork Time Management Creativity

Hobbies

Solving Puzzles Video Editing