QueryMind

Agentic Business Intelligence assistant — ask questions about CSV/Excel data in plain English.

Upload a spreadsheet, ask in natural language, and QueryMind autonomously runs Pandas/SQL analysis, builds charts, and returns an explained answer. Built to demonstrate agentic tool use end to end.

---

Highlights

• Agentic workflow: LangChain agent with analysis, SQL, charting, and column tools
• No-code BI: plain-English questions over uploaded CSV/Excel
• Follow-up memory: ask chart variants and drill-downs in the same session
• Auto insights: dataset summary, KPI strip, suggested questions
• Polished UI: dark glassmorphism frontend served by FastAPI

---

How it works

Upload file → /upload → Pandas + SQLite + schema injected into LLM
Ask question → /chat  → agent picks tools → code execution → answer (+ optional chart)

---

Tech stack

Layer	Technology
Backend	FastAPI, Uvicorn
Agent	LangChain tool-calling agent
LLM	Groq LLaMA 3.3 70B
Data	Pandas, SQLite
Charts	Matplotlib
Frontend	HTML, CSS, Vanilla JS

---

Quick start

Prerequisites: Python 3.10+, free Groq API key

git clone https://github.com/Znaxh/QueryMind.git
cd QueryMind

python -m venv venv
source venv/bin/activate   # Windows: venv\Scripts\activate
pip install -r requirements.txt

echo "GROQ_API_KEY=your_key_here" > .env
uvicorn api:app --reload --port 8000

Open http://localhost:8000

Optional legacy Streamlit UI:

streamlit run app.py

---

Example questions

Question	What happens
Summarize this dataset	Row/column/type overview
Which category has the highest profit?	Pandas groupby analysis
Show a bar chart of sales by region	Matplotlib chart output
What are the top 5 products by revenue?	SQL over SQLite
Now show that as a pie chart	Follow-up with conversation memory

---

API

Method	Endpoint	Description
GET	/	Frontend UI
POST	/upload	Upload CSV/Excel and initialize agent
POST	/chat	Ask a question, get answer + optional chart

---

Project structure

QueryMind/
├── api.py              # FastAPI app
├── app.py              # Streamlit UI (legacy)
├── agent/
│   ├── core.py         # Agent + Groq setup
│   ├── tools.py        # Analysis / SQL / chart tools
│   └── prompts.py
├── utils/
│   └── data_loader.py
└── static/
    └── index.html

---

Roadmap

• Multi-file joins
• PDF export of conversations
• Auth + saved sessions
• Live database connectors

---

Author

Anurag Pratap Singh · GitHub · Portfolio

License

MIT — see LICENSE.