🎯 Machine Learning Engineer with 1.10 years of hands-on experience in building and deploying end-to-end AI solutions
🔬 Specialized in Computer Vision, AutoML Systems, and MLOps
🚀 Passionate about creating real-world impact through AI in healthcare, agriculture, and accessibility
📊 Expert in model optimization, achieving 18% accuracy improvement and 30% training time reduction
🎓 B.E in Agriculture Engineering from Mahendra Engineering College
Currently: Building scalable ML pipelines at Aislyn Technologies Pvt Ltd
Always: Exploring cutting-edge AI technologies and contributing to open source
End-to-end Automated Machine Learning System
| Aspect | Details |
|---|---|
| Goal | Build an intelligent AutoML system that automates the entire ML pipeline from data upload to model deployment |
| Tech Stack | FastAPI, MLflow, Docker, Scikit-learn, Pandas, NumPy, Python |
| Key Features | • Automated data preprocessing & feature engineering • Multiple algorithm comparison (Random Forest, SVM, XGBoost) • Hyperparameter tuning & model selection • Real-time prediction REST APIs • MLflow experiment tracking & monitoring • Docker containerization for scalable deployment |
| Impact | Reduced model development time by 60%, enabled non-technical users to leverage ML capabilities |
AI-Powered Agricultural Intelligence
| Aspect | Details |
|---|---|
| Goal | Predict soil nutrient levels to optimize crop yield and fertilizer recommendations |
| Tech Stack | Python, Scikit-learn, XGBoost, Random Forest, SVR, Pandas, NumPy, Flask |
| Key Features | • Comprehensive feature engineering from raw soil data • Ensemble learning combining Random Forest, XGBoost, and SVM • Hyperparameter optimization using GridSearchCV • Robust inference pipeline with preprocessing consistency • REST API for real-time predictions |
| Impact | Achieved 92% accuracy in nutrient prediction, helping farmers reduce fertilizer costs by 25% |
Real-time Security Intelligence Platform
| Aspect | Details |
|---|---|
| Goal | Develop an intelligent surveillance system for automated threat detection and facial recognition |
| Tech Stack | YOLOv8, OpenCV, Face Recognition, CNN, Python, SQLite, FastAPI |
| Key Features | • Real-time object detection using YOLOv8 (people, vehicles, suspicious items) • Face recognition for known/unknown person identification • Automated alert generation for security threats • Event logging system with SQLite database • Live monitoring dashboard with video streaming • Intelligent image processing and frame analysis |
| Impact | Reduced manual monitoring effort by 85%, improved threat detection accuracy by 40% |
Aislyn Technologies Pvt Ltd, Bengaluru | May 2024 – March 2026
Key Achievements:
- ✅ Built and deployed end-to-end ML pipelines for classification and regression tasks
- ✅ Improved model accuracy by 18% through advanced feature engineering
- ✅ Reduced training time by 30% with hyperparameter optimization
- ✅ Developed production-ready REST APIs using FastAPI
- ✅ Implemented MLOps practices with MLflow and Docker
- ✅ Deployed scalable solutions handling 1000+ requests/minute
Technologies Used: Python, Scikit-learn, TensorFlow, FastAPI, MLflow, Docker, Pandas, NumPy, XGBoost, Git, VS Code, SQL
⭐ If you like what you see, consider starring some repositories! ⭐