Amit P

Senior Python Developer @Bizinc

Atlanta, GA, US

Led the development of end-to-end AI/ML solutions for real-time systems by designing data ingestion pipelines, training models, and deploying RESTful endpoints for model inference using FastAPI and Flask.Built and deployed deep learning models for image classification, object detection, and NLP tasks using TensorFlow, Keras, PyTorch, and scikit-learn, optimizing model performance through hyperparameter tuning and evaluation strategies.Developed ETL and data pipeline automation using Apache Airflow and custom Python scripts, ensuring continuous data flow for model training and predictions from diverse sources such as Kafka, Spark, and PostgreSQL.Integrated ML models into web-based platforms using Django and Flask, enabling real-time predictions and AI-powered insights for end-users through intuitive front-end interfaces.Designed and implemented deep learning pipelines using TFX (TensorFlow Extended) and integrated OpenTelemetry SDK for tracing, logging, and monitoring model performance.Created real-time information pipelines using Kafka + Storm + HDFS for high-throughput data processing and analytics.Built robust ML solutions for fraud detection, sentiment analysis, recommendation systems, and predictive maintenance, integrating them into production via containerized microservices.Implemented Celery-based task queues for async ML workflows such as batch scoring, model re-training, and long-running model evaluation.Developed and documented RESTful APIs for model serving and telemetry data collection from IoT devices using Psycopg2 and PostgreSQL.Collaborated with data scientists and domain experts to translate business requirements into scalable AI solutions and drove MLOps best practices.Deployed ML workloads on Kubernetes clusters using Docker and Helm, with model versioning and artifact tracking via MLflow.Utilized Snowflake, AWS Athena, and Spark SQL for data exploration, feature generation, and distributed training preparation.