Concept Sprint GPT
Published:
Fun project built together with a Grant Thornton partner. We used LLMs to simulate a Concept Sprint for a given problem. We then used the results to generate a report for the client.
Built during the earlier days of LLMs using OpenAI API. Desparate attempt to get structured output at a time when JSON output was not a thing.
Built using Langchain, Streamlit and OpenAI API.