Traveler

• 

  Traveler

Inspiration

As we worked as software devs in DR, we realized something: Devs need to be more active. We don’t use it to document, report to our PMO, or even care about testing; as long as the functionality is released, we’re good. That’s when we thought... What if we build something to cut the time we spend documenting and reporting? And Traveler was born.

An environment, an ecosystem that will allow us to track the coding process. This could include monitoring how frequently we submit our changes for the PMO to see, checking code changes so we could create documentation, or integrating with a logging tool to document errors and resolve bugs for future reference.

We think of Traveler as a partner in the empress of building systems, giving you the fun and the part that you actually want to do: code.

What it does

We use a CLI tool to submit changes to your code, which are processed and analyzed by our Ollama model. We keep your project in an environment that makes it easy to export and share with your teammates.

How we built it

Model Training: We trained our model using OLLAMA. Additionally, we utilized Lang Chain to interact with proficiency and reduce costs.

Client Interaction: For the client, we developed a Command Line Interface (CLI) for the terminal. This CLI enables users to log in and seamlessly redirect to the web page for further interaction.

Authentication and Database: We employed Auth0 and MongoDB for their indexing capabilities, which allow us to efficiently interact with data when utilizing our model.

Backend Development: In the backend, we leveraged Django to abstract complex serialization tasks and streamline the use of the back-office functionalities.

Challenges we ran into

first we tried to use the rest api from github but we run a lot of problem specific with permissions and credentials so we looked another alternative to recolect data using our cli and it works better than we expected

Accomplishments that we're proud of

We are proud of the entire proyect and all of the integrations that we made in 24 hours, the design was on point and the user experience with our features works perfect

What we learned

We learn about how to use efficiently the new ia tecnology how we can chain method in langchain and output really good responses and for free

What's next for Traveler

We strongly believe that this project holds immense potential. Our goal is to create the perfect experience for developers who often find themselves exhausted while working in large codebases. Additionally, we aim to provide stakeholders with a platform that facilitates access to key information during crucial moments in the project lifecycle.

Built With

• auth0
• django
• mongodb
• ollama
• python
• react
• tailwind