Intellens

Inspiration

During the summer, our team members worked as software engineering interns. We often had to submit architecture diagrams and conduct code reviews. We realized that understanding large and unfamiliar codebases was extremely time-consuming. It took many hours to determine how different files interacted, which services were utilized, and what technologies powered each component. This experience inspired us to create Intellens, which stands for Intelligent Lens. Intellens is a tool designed to make understanding complex codebases much easier. It automatically analyzes code, visualizes the architecture, and helps developers quickly understand how everything works together.

What it does

Intellens serves as an intelligent lens that analyzes any uploaded project in depth. When a user uploads a project as a ZIP file, Intellens automatically processes it to provide a comprehensive understanding of that specific codebase. It begins by generating a detailed Project Summary that includes an overview of the uploaded project, the programming languages used, and the services detected within it. Then it builds a System Architecture Diagram that visualizes how these languages and services interact with each other inside the project. Intellens also constructs an Infrastructure Diagram that shows how the languages and internal cloud services connect within the project’s architecture. Following this, it creates a Component Diagram that organizes the backend, frontend, and configuration layers, helping users clearly see how different parts of the project communicate and function together. Finally, Intellens provides a preview of the frontend interface and generates a structured README file that summarizes the entire project.

How we built it

Intellens is built using FastAPI (Python backend) and vanilla HTML/JavaScript (frontend). The backend uses multi-language parsers to analyze uploaded ZIP projects, detecting technologies through file extensions and content scanning. We used Claude 3 generates intelligent descriptions and summaries. The system creates interactive workflow diagrams using custom JavaScript renderers, supports Terraform/AWS infrastructure analysis, and generates Mermaid diagram syntax.

Challenges we ran into

One of the main challenges we faced was designing the frontend experience. We wanted to make Intellens not only functional but also engaging, so we initially explored creating an interactive “Lego-style” interface where each block represented a system component. We experimented with different Lego-like diagrams, drag-and-drop interactions, and visual connections between modules to make learning and code understanding more enjoyable. Another challenge was building the infrastructure layer that visualizes internal cloud components and how they interconnect. Ensuring accuracy while maintaining clarity in these visualizations required multiple design iterations. We also had to optimize the backend to make the AI-powered code analysis faster and more efficient, especially when processing large projects.

Accomplishments that we're proud of

We are proud of building a complete end-to-end system that can intelligently analyze any uploaded project and generate meaningful insights within minutes. One of our biggest achievements was successfully integrating AWS Bedrock to perform accurate and context-aware code analysis across multiple programming languages. We are also proud of designing a clean and interactive frontend that visualizes architecture diagrams, workflows, and project layers in real time. Our key accomplishment is that Intellens makes code understanding simpler, faster, and more accessible for developers everywhere.

What we learned

Through building Intellens, we learned how to bring together AI, cloud computing, and software design into one cohesive system. We gained a deeper understanding of how to integrate AWS Bedrock for intelligent analysis and how to design APIs that efficiently handle large and complex projects.

What's next for Intellens

Next, we plan to enhance Intellens by integrating it with project management tools such as Jira. This will allow users to automatically generate future tasks based on the uploaded project, assign responsibilities, and plan the next steps efficiently. In addition, Intellens will continue to improve automated deliverables, including architecture diagrams, infrastructure maps, component overviews, frontend previews, and comprehensive README files.