From the article, they did not use fine-tuning. This is semantic search + GPT-3 to provide human-like answers.
The semantic search approach seems to focus the answers better than fine-tuning; at the cost of preloading the prompt with a lot of tokens, but with the benefit of a more constrained response.