● Built and deployed personalized recommendation systems using TensorFlow for e-commerce platforms. ● Developed real-time video surveillance models with PyTorch for object detection and anomaly detection. ● Deployed scalable ML solutions on Google Cloud ML Engine for high-performance data processing. ● Automated training pipelines using AWS SageMaker, reducing development time by 30%. ● Designed real-time recommendation APIs using FastAPI and TensorFlow Serving, reducing response latency. ● Implemented distributed model training pipelines on Apache Spark for large-scale datasets. ● Processed and optimized large datasets with Pandas and NumPy for machine learning pipelines. ● Applied OpenCV and DeepFace for facial recognition and emotion detection in video analytics. ● Improved ML inference scalability by implementing serverless functions with AWS Lambda. ● Accelerated distributed data processing workflows with Apache Spark and Dask. ● Developed custom transformers in PyTorch for sequence-to-sequence tasks. ● Developed keyword extraction solutions with NLTK and SpaCy for NLP-based search systems. ● Built multi-label text classifiers with Scikit-Learn for automated categorization. ● Implemented Prophet and Statsmodels for time-series forecasting. ● Developed explainable AI (XAI) models to improve model transparency and interpretability for stakeholders. ● Integrated MLflow and Docker for experiment tracking and model deployment. ● Integrated TensorFlow Recommenders to enhance personalized product recommendations. ● Implemented adversarial training techniques to improve model robustness against data perturbations. ● Trained CNN models in Keras for sentiment analysis in NLP tasks. Tech Stack: TensorFlow, PyTorch, FastAPI, OpenCV, AWS SageMaker, Google Cloud ML Engine, Apache Spark, Docker, Kubernetes, Pandas, NumPy, PostgreSQL.

● Built product recommendation engines with TensorFlow, increasing cross-sell conversions by 20%. ● Developed real-time sales forecasting models using Prophet for predicting seasonal demand. ● Applied OpenCV for video capture analytics and product tracking in retail environments. ● Designed churn prediction models with XGBoost, improving customer retention strategies. ● Built automated inventory models that improved stock-level accuracy and reduced overstock by 15%. ● Created Streamlit dashboards to visualize customer engagement and sales trends. ● Integrated graph-based fraud detection with PyTorch Geometric for anomaly detection. ● Conducted time series analysis with Prophet and Statsmodels for forecasting financial market trends. ● Designed customer segmentation strategies using KMeans and hierarchical clustering for targeted marketing campaigns. ● Developed anomaly detection systems in financial datasets using AutoML and CatBoost. ● Streamlined MLOps workflows with GitHub Actions and MLflow for continuous model integration and deployment. ● Preprocessed large datasets with Apache Spark and Dask for model training and analysis. ● Developed voice recognition systems with Google Speech-to-Text APIs for hands-free mobile control. ● Built A/B testing frameworks for evaluating personalization strategies, boosting engagement by 25%. ● Built and optimized data dashboards using Streamlit and Dash for visualizing model results. Tech Stack: TensorFlow, PyTorch, XGBoost, Prophet, Statsmodels, OpenCV, Apache Spark, PyTorch Geometric, Streamlit, PostgreSQL, Docker, Kubernetes.

● Developed scalable backend services with Django and Flask for AI model deployment and data integration. ● Developed NLP models with SpaCy and NLTK for document classification in contract automation. ● Designed cash flow prediction models with Prophet, improving spend analysis accuracy. ● Applied transfer learning techniques with pre-trained models to accelerate NLP tasks. ● Built contract risk prediction models using XGBoost, identifying potential supplier risks with 85% accuracy. ● Optimized backend services for scalable model deployment using Gunicorn and NGINX. ● Built named entity recognition (NER) models for financial documents to extract key supplier details. ● Built real-time NLP pipelines with NLTK and SpaCy, improving text analysis and named entity recognition accuracy. ● Applied biometric authentication using FaceNet and OpenCV for identity verification. ● Designed and deployed recommendation systems with Scikit-Learn for internal knowledge management. ● Built time series forecasting models using Prophet and Statsmodels for sales trend prediction. ● Implemented secure APIs with Django REST Framework, ensuring seamless integration with third-party services. Tech Stack: Django, Flask, SpaCy, NLTK, Scikit-Learn, DeepFace, OpenCV, Prophet, PostgreSQL, MongoDB, Docker.