Technical Architect
CurrentResearch Leader and AI Data Scientist/ArchitectArts: Internal Revenue Service AI ProjectWeb platform and mobile app for digitization of documents from A.D.E. folders.Engineering and prototyping of 5 AI models for object detection of folders and handwritten textpresent on the folders, recognition of written text (printed and handwritten) in edge mode.Design and optimization of classification and object detection models for operation on edge-type devicesand embedded operation.Technical architecture design on prototype and for production environment.Design and implementation of DAGs and execution pipeline of AI models in production environment.Design and implementation in production environment of MLOps best practices for managing data lifecycles andof the implemented modelsDesign, Training, Inference and Tuning of neural network architectures for recognized text classification modelsTechnologies: Tensoflow, Pytorch, Airflow, MLflow, BentoML, Python, Docke Composition, Kubernetes, Pascal voc writer,opencv, matplotlib, scikit-learn, tflite-model-maker, springboot microservices, mongodb.