\section{Inspiration}

College students often struggle to turn scattered academic information into a clear plan. Managing deadlines, assignments, and study schedules across multiple platforms can be overwhelming and time-consuming.

This inspired \textbf{Tactica AI} --- a personal academic strategist that transforms academic information into clear, actionable plans. Using each student's academic context, Tactica AI generates personalized semester roadmaps, weekly priorities, exam preparation plans, and academic recommendations.

Rather than being just another calendar, task manager, or generic chatbot, Tactica AI brings academic information, planning, studying, and decision-making together in one intelligent system.

\section{What We Built}

Tactica AI is an AI-powered personal academic strategist that transforms scattered academic information into personalized, actionable plans.

Students upload syllabi and course materials, and Tactica AI extracts key information such as deadlines, exams, assignments, and course topics.

\begin{center} \textbf{Academic Documents $\rightarrow$ Structured Data $\rightarrow$ AI Reasoning $\rightarrow$ Actionable Strategy} \end{center}

Using that structured context, Tactica AI can generate personalized semester roadmaps, weekly priorities, exam preparation plans, and academic recommendations.

Instead of being another calendar, task manager, or generic chatbot, Tactica AI is designed to connect \textbf{academic information, planning, studying, and decision-making} in one intelligent system.

\section{How We Built It}

We designed Tactica AI as a document-centered AI system, not simply a traditional productivity app with an LLM added on top.

\begin{itemize} \item \textbf{Frontend:} Next.js, React, TypeScript, and Tailwind CSS \item \textbf{Backend:} FastAPI and Python \item \textbf{Database:} PostgreSQL \item \textbf{Vector Retrieval:} pgvector \item \textbf{AI:} Google Gemini \end{itemize}

We use \textbf{PostgreSQL} to store structured academic data and design the system to support vector-based retrieval through \textbf{pgvector}.

\textbf{Google Gemini} serves as the main reasoning and generation layer.

Rather than sending every user question directly to an LLM, our architecture follows \textbf{Retrieval-Augmented Generation (RAG)} principles. Academic documents are processed, relevant information is retrieved, and that context is provided to Gemini before it generates a response.

This allows the AI to reason about the student's actual courses and academic materials instead of relying only on general model knowledge.

The broader architecture also supports background workflows for time-sensitive features such as study reminders, exam countdowns, recurring recommendations, and progress tracking.

\section{Challenges We Ran Into}

One of our biggest challenges was the steep learning curve. We had to familiarize ourselves with many aspects of building a full-stack AI product, and much of it was completely new to us. We had to learn as we built --- from frontend and backend development to databases, APIs, authentication, cloud deployment, and AI integration.

We also realized that building an AI-powered product is not as simple as calling an LLM API. Tactica AI needs to process academic documents, structure and retrieve relevant information, and generate reliable recommendations. Since inaccurate deadlines or academic information could affect students' real-world decisions, AI reliability and validation became another major challenge.

\section{What We Learned}

Building Tactica AI taught us that creating an AI-native product involves much more than connecting an application to an LLM API.

We learned about document processing, structured data modeling, retrieval systems, RAG, backend architecture, validation, asynchronous workflows, authentication, and collaborative software development.

More importantly, we learned that the value of AI does not come from generating more information.

\begin{center} \textit{It comes from turning existing information into useful decisions and actions.} \end{center}

\section{What's Next for Tactica AI}

Our next step is to make Tactica AI a more adaptive and proactive academic strategist. We plan to improve personalized study recommendations, academic risk detection, and recovery plans so the system can adjust as a student's semester changes.

We also want to expand degree and course planning, allowing students to build long-term academic roadmaps based on their degree requirements, prerequisites, and graduation goals.

Ultimately, our goal is to grow Tactica AI beyond an academic planning tool into an intelligent companion that supports students throughout their entire college journey.

Built With

Share this project:

Updates