The Professor

Inspiration

As students, we realised that learning while coding is one of the biggest hurdles in mastering Computer Science. Most of us have spent nights staring at lines of code we didn’t understand which was not because we couldn’t code, but because we didn’t truly grasp the “why” behind what we were building. Office hours are limited. Professors can’t be everywhere. And while tools like Copilot can complete your code, they rarely teach you the theory behind it. That gap between completion and comprehension is exactly what inspired us to build The Professor: an AI-powered academic mentor who explains your code the way a real professor would. We wanted to reimagine what learning feels like which is not just faster, but also deeper.

How we built it

~Frontend & UX: We crafted an intuitive interface, ensuring students could “Ask the Professor” effortlessly whether they wanted a line-by-line breakdown, a Big-O analysis, or a beginner-friendly explanation. ~VS Code Extension: We implemented the core functionality using TypeScript and VS Code’s Extension API, creating custom commands and context-aware menus that blend naturally into the coding workflow. ~AI Integration: We connected to the Google Gemini API, building prompts that adapt to different levels of understanding from “Explain like I’m 5” to “Explain the algorithmic complexity.” ~Team Collaboration: Each member contributed to different modules like documentation, testing, UX design, and API logic merging our skills into a single cohesive product.

Languages: TypeScript, JavaScript, HTML, CSS Frameworks & Tools: Node.js, VS Code API AI & APIs: Google Gemini API for academic explanations Version Control: Git & GitHub (SSH-based setup for collaboration)

Challenges we ran into

Building The Professor wasn’t easy and that’s what made it meaningful. ~Balancing AI and Accuracy: Designing prompts that didn’t just give surface-level answers but true academic insight took multiple iterations. We learned how to guide the model toward educational depth rather than code autocompletion. ~ VS Code Integration: Debugging the extension environment was challenging. The extension had to respond instantly within the editor where there must be no pop-ups, no latency, no clutter. ~API Key Management: Ensuring secure and persistent API keys while maintaining a smooth user experience was trickier than we expected. ~Time Pressure: As students with limited time (and sleep), managing deadlines while coding late into the night taught us the real meaning of collaboration and perseverance.

Accomplishments that we're proud of

One of our biggest technical wins was making Gemini which is naturally stateless, context-aware. By default, every Gemini API call starts fresh with no memory, but we engineered a custom “context bundling” system inside our VS Code extension. It dynamically scans open files, detects related code, merges them into a single contextual snapshot, and uploads it through the Gemini Files API before each query. This means the AI doesn’t just see a few selected lines but it also understands the entire project structure, relationships, and logic flow. This resulted in exponentially smarter answers resulting in explanations becoming richer, more precise, and unbelievably relevant to the user’s code. We essentially taught Gemini to think like a professor who remembers the whole class, not just one question.

What we learned

Building The Professor was an immersive journey into both education and engineering. We explored how AI can be used not just as a productivity tool, but as a teaching companion. Throughout this project, we learned to: ~Integrate LLM APIs into a VS Code Extension. ~Understand prompt engineering and the nuances of generating pedagogical explanations. ~Work collaboratively using TypeScript, JavaScript, and Node.js for extension logic. ~Optimize context handling and data flow within VS Code’s API environment. Each of us focused on areas that matched our strengths like UI/UX, backend logic, API integrations, and prompt design where we learnt to rely on each other’s expertise.

What's next for The Professor

Our next step is to evolve The Professor from a hackathon prototype into a full-scale academic learning companion. We envision a version that combines chat-based explanations with real-time academic guidance meaning its a conversational AI that doesn’t just answer coding questions, but teaches concepts interactively, adapting to every learner’s pace. The current VS Code extension is only the beginning. We plan to integrate The Professor into a standalone chatbot interface, where students can: ~Paste or upload their code and receive interactive explanations. ~Ask for conceptual help (“Explain recursion with an example”) or algorithmic insight (“What’s the time complexity here?”). ~Save chats, track progress, and revisit past lessons like digital office hours. This will make The Professor accessible beyond the editor, opening it to anyone learning computer science from high school beginners to college students struggling through algorithms and data structures. We also believe education tools should be affordable, not exclusive. That’s why our next goal is to release The Professor at a minimal price point, just enough to sustain server and API costs while keeping it accessible to everyone. We also aim to partner with universities and educational institutions to bring The Professor directly to classrooms and learning centres where it can be offered to students at a significantly reduced or subsidised rate through campus-wide collaborations. By doing this, we can ensure that The Professor remains a sustainable project that helps students learn deeply, without financial barriers. We deeply believe learning to code shouldn’t be about just getting the right answer, it should be about understanding why it’s right.