Applied AI/ML Engineer | Production ML | MLOps | Forecasting | Ranking & Retrieval

I am an Applied ML Scientist and Ph.D. researcher focused on building machine learning systems from experimentation through production deployment.

My work spans:

• Production ML and MLOps
• Time-series forecasting
• Computer vision
• Anomaly and fault detection
• Ranking, retrieval, and recommendation systems
• Generative AI and LLM-based applications

I work primarily with Python, PyTorch, TensorFlow, scikit-learn, MLflow, Docker, AWS, and FastAPI, with experience developing reproducible ML pipelines, deploying inference services, and evaluating models against both technical and operational metrics.

In development

Production-oriented retail ML platform combining:

• semantic product retrieval
• recommendation and ranking models
• user/product feature pipelines
• offline ranking evaluation
• model serving APIs
• MLflow experiment tracking
• CI/CD and containerized deployment
• model monitoring and retraining workflows

Focus: Retrieval · Ranking · Recommendation · NLP · MLOps

Repository

End-to-end hybrid ML/DL system for pulmonary disease classification from CT scans.

• CNN-based feature extraction
• classical ML classification
• MLflow experiment tracking
• DVC data/version management
• Dockerized inference service
• GitHub Actions CI/CD
• AWS deployment

• 96% macro F1-score on the evaluation dataset

Focus: Computer Vision · Deep Learning · MLOps · AWS

Repository

Predictive-maintenance pipeline for early fault detection using multivariate turbine sensor data.

• unsupervised autoencoder pretraining
• transfer learning
• imbalanced classification
• SHAP-based model interpretation
• modular ingestion, transformation, training, and evaluation pipelines
• MLflow experiment tracking

Focus: Anomaly Detection · Deep Learning · Predictive Maintenance

Repository

Cloud-native ML workflow demonstrating:

• managed training on AWS SageMaker
• S3-based data and artifact storage
• endpoint deployment
• model evaluation
• SHAP-based explainability

Focus: AWS SageMaker · Cloud ML · Model Deployment

Python · PyTorch · TensorFlow · scikit-learn · Keras · Hugging Face
Forecasting · Computer Vision · Anomaly Detection · Ranking · Retrieval · Regression · Classification

MLflow · DVC · Docker · FastAPI · Flask · GitHub Actions · CI/CD · Kubernetes
Model versioning · Experiment tracking · Inference APIs · Model monitoring · A/B testing

SQL · Pandas · NumPy · SciPy · PySpark · ETL pipelines

AWS · SageMaker · S3 · GCP

Prometheus · Grafana · Kubernetes · Docker

I am currently deepening my work in:

• production-grade ranking and recommendation systems
• semantic retrieval and product search
• scalable time-series forecasting
• model serving and inference optimization
• ML observability and drift detection
• canary, shadow, and A/B model deployment strategies

• Model training and evaluation
• Experiment tracking
• Model serving
• CI/CD
• Cloud deployment
• Monitoring and retraining

• Time-series forecasting
• Computer vision
• Anomaly detection
• Ranking and retrieval
• Generative AI

• AWS / SageMaker
• Docker
• Kubernetes
• MLflow
• FastAPI