Drawer

Research Workflows for Carbon-Optimal AI Inference Scheduling

For this project, I applied the ChatGPT workflows from the masterclass to streamline the research phases for a system designed to optimize AI inference scheduling.

  • Brainstorming & Ideation: Used structured prompting to generate foundational concepts for a grid-aware LLM energy agent.

  • Literature Review: Summarized dense academic papers regarding machine learning and predictive analytics to quickly identify gaps in current carbon-reducing methodologies.

  • Outline Generation: Created a comprehensive framework and structural foundation for the initial proposal.

Implementing these workflows drastically reduced the time spent synthesizing preliminary research and provided clear direction for the project's development.