Exploring function calling, structured output and agents capabilities of current llms.

Plan:

• Test current function calling capabilities.
• Test current agent strategies and test new ones.
• Create an eval for different tasks to measure the reliability of different llms and implementations.