Los Caris

Inspiration

We were inspired by the challenge of onboarding new developers efficiently while ensuring they adhere to the company’s coding standards and best practices. We wanted to create a tool that would automate and simplify this process, making it easier for new hires to get up to speed with company-specific code guidelines while receiving immediate feedback.

What it does

Our platform automates the onboarding process by extracting coding guidelines directly from the company’s GitHub repository. It generates tailored programming exercises using generative AI, provides static code analysis, and offers real-time feedback. This ensures new developers quickly align with company standards and best practices from day one.

How we built it

We built the platform by integrating several key technologies. We used MongoDB for data storage, Python for backend logic, and QT/QML to create the user interface. We also implemented a chatbot using the OpenAI API and Huggingface models to handle questions about coding guidelines and best practices. To power the code comparison and feedback system, we integrated Pinecone for efficient search and similarity matching. Together, these technologies allowed us to automate the onboarding process and provide real-time, AI-driven feedback to new developers.

Challenges we ran into

One of the main challenges was integrating dynamic code analysis with static analysis for more personalized feedback. Additionally, implementing a functional web interface was challenging due to time constraints, which limited some of our planned features.

Accomplishments that we're proud of

We successfully automated a significant part of the onboarding process, from pulling guidelines from GitHub to creating customized exercises. We're proud of the platform's ability to analyze code and provide actionable feedback, making it a useful tool for any company's onboarding process.

What we learned

We learned the importance of balancing automated feedback with human-readable insights. The process of integrating static code analysis taught us how to streamline code quality checks, and we gained valuable experience working with generative AI models for generating and verifying code based on a restrained knowledge base.

What's next for Los Caris

Next, we plan to implement dynamic code analysis to provide more comprehensive feedback and better align with real-time development scenarios. Additionally, we're working on integrating the platform into a web interface for a smoother user experience, which we couldn’t finish due to time limitations.