DEV Community

Shaik Bibi Jahera
Shaik Bibi Jahera

Posted on

How I Built an AI Study Planner in 3 Days Using Coral SQL, Groq AI and Streamlit

#ai #hackathon #beginners #python

The Problem That Started It All
As a CSE student, I waste time every morning figuring out what to study. I open my Google Calendar, then switch to Notion, then try to piece it all together manually.
It's exhausting.
So when I saw the Pirates of the Coral-bean Hackathon by WeMakeDevs, I had one goal — build ONE tool that reads my schedule and notes and just tells me what to study. Automatically.
That tool became StudyMate. 📚

What is StudyMate?
StudyMate is an AI-powered study assistant that:

Reads your Google Calendar events and deadlines
Reads your Notion notes and assignments
Uses Groq AI (Llama 3.3 70B) to generate a personalized 7-day study plan
Displays everything in a beautiful Streamlit web UI

One click. Everything sorted.

The Magic — Coral SQL
Here's the part that genuinely blew my mind.
Normally, connecting Google Calendar and Notion would require:

Separate API integrations for each service
Handling OAuth for each one
Parsing different response formats
Writing glue code to combine everything

Hours of work.
With Coral, I just wrote:
sqlSELECT summary, start_date FROM google_calendar.events LIMIT 20
SELECT url FROM notion.search LIMIT 20
That's it. Two SQL queries. Two data sources. Zero extra code.
Coral turns any app, API, or file into a SQL table. You query it like a database. No extra setup needed.

How It Works
Google Calendar + Notion
↓
Coral SQL
↓
Python Agent
↓
Groq (Llama 3.3)
↓
📚 Personalized Study Plan

*Step-by-Step: How I Built It
*Step 1 — Install Coral
Download from withcoral.com and add to PATH.
bashcoral --version
Step 2 — Connect Data Sources
bashcoral source add --interactive google_calendar
coral source add --interactive notion
Follow the OAuth prompts for Google Calendar and paste your Notion API token.
Step 3 — Test Your Queries
bashcoral sql "SELECT summary, start_date FROM google_calendar.events LIMIT 5"
coral sql "SELECT url FROM notion.search LIMIT 10"
Seeing your real data come through SQL feels like magic! ✨
Step 4 — Build the Agent
pythonimport subprocess
from groq import Groq

def query_coral(sql):
result = subprocess.run(
["coral", "sql", "--format", "json", sql],
capture_output=True, text=True
)
return result.stdout

def get_study_plan():
events = query_coral("SELECT summary, start_date FROM google_calendar.events LIMIT 20")
notion_pages = query_coral("SELECT url FROM notion.search LIMIT 20")

client = Groq(api_key="YOUR_GROQ_API_KEY")

response = client.chat.completions.create(
    model="llama-3.3-70b-versatile",
    messages=[{
        "role": "user",
        "content": f"""You are StudyMate, a smart student assistant.
Enter fullscreen mode Exit fullscreen mode

Google Calendar events: {events}
Notion pages: {notion_pages}

  1. Identify exams, assignments, study sessions
  2. Create a prioritized study plan for next 7 days
  3. Give 3 actionable tips for today Use emojis, be friendly and motivating!""" }] ) return response.choices[0].message.content Step 5 — Add a Beautiful Web UI with Streamlit pythonimport streamlit as st

st.title("📚 StudyMate")
st.subheader("Your Personal AI Study Assistant")

if st.button("🚀 Generate My Study Plan"):
with st.spinner("Fetching your data..."):
plan = get_study_plan()
st.markdown(plan)
Step 6 — Run It!
bashstreamlit run app.py

What I Learned
→ Coral is genuinely game-changing for agents
The ability to query any app as SQL removes 90% of the boring integration work. I can focus on building the actual agent logic.
→ Building an AI agent is simpler than I thought
With the right tools (Coral + Groq), building a working AI agent took less than a day.
→ Ship fast, polish later
I built a working version in Day 1 and spent Day 2 polishing the UI and README. Done is better than perfect.
→ Hackathons are the best way to learn
Nothing beats building something real under deadline pressure!

Try It Yourself!
🔗 GitHub: https://github.com/jahera-shaik/StudyMate
🎥 Demo: https://www.loom.com/share/b1f166c7a827462caa1ba51ef298c74e
Quick Setup:
bash# Install Coral

Download from withcoral.com

Connect sources

coral source add --interactive google_calendar
coral source add --interactive notion

Install dependencies

pip install streamlit groq python-dotenv

Run

streamlit run app.py

Final Thoughts
StudyMate started as a hackathon project but it's now something I actually use every day.
The biggest lesson? The best tools make hard things feel obvious. Coral made connecting two completely different apps feel as simple as writing two SQL queries.
If you're a student, try building something that solves YOUR problem. You'll be amazed what you can ship in just 2 days! 💪

Built for the Pirates of the Coral-bean Hackathon by WeMakeDevs
By Bibi Jahera Shaik

Top comments (0)