2nd Guess

Inspiration

AI guesses what you want instead of asking the right questions in the right way. We created a skeleton for developers to implement a smart chat that improves the quality of the information provided, and once the user is satisfied with the completeness of the chat, it can be documented in accordance with industry best practices.

What it does

It enables a voice and text chat that interacts with the user, collecting data and replying with a summary and following up questions, enhancing the quality of the conversation and documentation.

How we built it

Through the opportunity to use Kiro, we heavily applied the tool from the beginning to the end. Not only generating specs, but also coding and testing. Moreover, it created monitoring tools and, through different CLI integrations, eased the code management and deployment.

Challenges we ran into

In some cases, we had to be more precise in our descriptions, and some errors would have been avoided had Kiro considered some constraints while building the code. Also, the interface is not very strong and requires more time to work.

Accomplishments that we're proud of

With this skeleton we generated two applications that, more than improve the experience of chatting with AI, consolidate focused knowledge and allows, system managers and developers understand how the LLM behaves and can act to improve the outcomes.

Built With

• kiro
• typescript