This research explores how Large Language Models can transform the complex lifecycle of deploying machine learning on resource-constrained IoT devices (TinyML).

• Introduces a framework for LLM-assisted TinyML development across the entire lifecycle
• Evaluates LLMs' capabilities in code generation, optimization, and debugging for embedded systems
• Identifies opportunities where LLMs excel and limitations requiring human expertise
• Provides practical demonstrations of LLM assistance in real TinyML implementations

For engineering teams, this research offers a practical roadmap to accelerate embedded AI development while reducing expertise barriers, potentially democratizing TinyML implementation across industries where edge computing is critical.

Consolidating TinyML Lifecycle with Large Language Models: Reality, Illusion, or Opportunity?