Data Scientist | MLOps Specialist | Specializing in Building Scalable AI Systems
- 🛠️ MLOps Engineer passionate about automating and scaling machine learning solutions.
- 🧠 Data Scientist experienced in Machine Learning, Deep Learning, NLP, and Computer Vision.
- ☁️ Skilled in Cloud Deployments (AWS EC2, S3, EKS, ECR) and CI/CD Automation (GitHub Actions).
- 🛡️ Focused on building production-grade, maintainable, and scalable AI systems.
- 🔹 End-to-End MLOps Pipelines
- 🔹 Machine Learning and Deep Learning Solutions
- 🔹 Natural Language Processing (NLP) and Computer Vision
- 🔹 Experiment Tracking (MLflow), Model Versioning (DVC), Model Registry
- 🔹 CI/CD Automation (GitHub Actions)
- 🔹 Docker, Kubernetes (AWS EKS) Deployment
- 🔹 AWS Cloud Services (EC2, EKS, S3, ECR, IAM)
- 🔹 Monitoring and Observability (Prometheus, Grafana)
- 🔹 Scalable ML Systems Architecture
- 🐍 Python | SQL | Bash
- 🧪 MLflow | DVC | Airflow
- 📦 Docker | Kubernetes (AWS EKS)
- ☁️ AWS (S3, EC2, ECR, IAM, EKS)
- ⚙️ Git | GitHub Actions | CI/CD Pipelines
- 🛠️ Prometheus | Grafana for Monitoring
- Built a production-ready NLP MLOps pipeline for sentiment analysis.
- Integrated DVC for data versioning, MLflow for experiment tracking.
- Deployed as a Dockerized microservice on AWS EKS with CI/CD via GitHub Actions.
- Real-time monitoring and alerting with Prometheus and Grafana.
- Developed an end-to-end MLOps solution to predict vehicle insurance responses.
- Achieved a 23.5% improvement in F1-score through model optimization.
- Deployed pipelines on AWS, utilizing Docker, CI/CD workflows, and MongoDB.
- Engineered a scalable machine learning pipeline for visa approval prediction.
- Achieved 95% model accuracy, deployed with Dockerized CI/CD workflows.
- Data storage and retrieval managed via MongoDB.
- Designed a modular ML pipeline to predict food delivery times.
- Implemented advanced regression models including XGBoost and Random Forest.
- Ensured robust data validation, logging, and custom exception handling.
💬 Feel free to reach out if you'd like to collaborate, discuss potential opportunities, or just connect: 📧 Email: [email protected]
Thanks for stopping by! 🚀