As part of the challenge, we developed and tested a pipeline to enable LLMs to perform API calls. An LLM is provided a toolbox of APIs and through prompt engineering and finetuning, it is able to make API calls whenever necessary. This method is also robust to new APIs.