praveen pitawala

Associate Machine Learning Engineer

CDAZZDEV•  April 2024 - Present

Developed a system to identify chemical engineering components in Process Flow Diagrams (PFD), and calculate carbon emissions using GPT-4 integration. Achieved high detection accuracy and significantly improved data processing workflows. Developed drawing interface for manipulate PFD Diagrams using JavaScript. Spearheaded the development of a backend code generation system that seamlessly integrated with the provided frontend code. Conducted extensive research utilizing LLMA2 and Mistral models, incorporating OpenAI capabilities to streamline development processes and enhance code efficiency.

Machine Learning Engineer Intern

CDAZZDEV•  September 2023 - March 2024

Spearheaded the development of a system to identify components in chemical process flow diagrams. Utilizing YOLO object detection techniques, gathered and pre-processed images, trained models, and evaluated their performance. Proficient in OpenAI tools, Hugging Face, AWS, GitHub, HTML/CSS/JavaScript, and image processing libraries like OpenCV and Pillow. Implemented an interface using the Flask framework for seamless integration.