StudyFlow — AI Student Assistant

Inspiration

Students often use separate tools for assignments, deadlines, study planning, and learning questions. We wanted to create one conversational experience where a student can simply ask for what they need and get an actionable response.

For example:

“Show my study tasks.”

“Add a task to finish recursion.”

“Make a 3-hour study plan for Accounts and Economics.”

“Explain depreciation in simple terms.”

“Complete task 2.”

What it does

StudyFlow is a conversational AI-style student productivity assistant that helps students organize academic work and learn concepts through natural-language requests.

It provides six core capabilities:

Add study tasks with subjects and deadlines

View saved study tasks and completion status

Mark study tasks as completed

Create simple study plans from available study hours

Explain academic topics at a beginner-friendly level

Provide study advice for a subject

Tasks are stored locally using SQLite, so completed actions remain available in the student's task list

How we built it

StudyFlow uses a simple architecture:

Web UI → Web API → Assistant → MCP Server → SQLite

The web interface provides the conversational student experience. The assistant interprets natural-language requests and connects them to application functionality. The MCP server exposes the student productivity capabilities as tools over Streamable HTTP, while SQLite provides persistent task storage.

The six MCP tools are:

add_task

list_tasks

complete_study_task

create_study_plan

explain_topic

get_study_advice

Challenges we ran into

One of the main challenges was understanding how MCP tools, client sessions, and Streamable HTTP fit together. We also wanted the assistant to remain useful without depending on a paid AI API, so we built a lightweight local assistant layer and connected it to the MCP tools and SQLite database.

Accomplishments that we're proud of

We are proud that we built a complete working prototype from the ground up during the hackathon.

The project combines a conversational student interface, persistent SQLite task storage, an MCP server with six usable tools, and Streamable HTTP into one working system.

We are especially proud that a natural-language request such as “Make a 3-hour study plan for Accounts and Economics” can be converted into a structured action and displayed immediately in the StudyFlow interface.

We also built the project without depending on a paid AI API, keeping the prototype lightweight and locally runnable.

What we learned

One of the main challenges was understanding how MCP tools, client sessions, and Streamable HTTP fit together. We also wanted the assistant to remain useful without depending on a paid AI API, so we built a lightweight local assistant layer and connected it to the MCP tools and SQLite database.

What We Learned

We learned how to build an MCP-based application, expose functionality as tools, connect a conversational interface to those tools, and persist user actions with SQLite.

The project also helped us understand how a conversational interface can turn simple student requests into concrete actions.

What's next for StudyFlow — AI Student Assistant

Our next goal is to make StudyFlow more personalized and useful for real students.

We would like to add calendar integration, exam countdowns, personalized study schedules, progress tracking, reminders, voice interaction, and adaptive study plans based on a student's completed tasks and available time.

We also want to expand the learning capabilities so StudyFlow can provide richer explanations, practice questions, revision plans, and subject-specific study support.

Our long-term vision is to make StudyFlow a conversational academic companion that helps students plan, learn, and stay organized through a single natural-language interface.

Built With

• css
• html
• http
• javascript
• mcp
• python
• sqlite
• streamable