README.md

metadata

title: DataMind Agent
emoji: 🧠
colorFrom: purple
colorTo: green
sdk: streamlit
sdk_version: 1.55.0
app_file: app.py
pinned: false

🧠 DataMind Agent

Multi-LLM AI Data Analyst — LangChain + Streamlit

Upload any data file (CSV, Excel, JSON) and chat with your data using natural language. Choose from 7 AI providers with the latest models — use whichever API key you already have.

---

🤖 Supported Providers & Models

🔵 Google Gemini

Series	Models
Gemini 3 (Latest)	gemini-3.1-pro-preview · gemini-3-flash-preview · gemini-3.1-flash-lite-preview
Gemini 2.5	gemini-2.5-pro · gemini-2.5-flash · gemini-2.5-flash-lite
Gemini 2.0 (Legacy)	gemini-2.0-flash · gemini-2.0-flash-lite
Gemini 1.5 (Legacy)	gemini-1.5-pro-002 · gemini-1.5-flash-002

🔑 Get API key

---

🟢 OpenAI GPT

Series	Models
GPT-5 (Latest)	gpt-5.4-pro · gpt-5.4 · gpt-5.3-instant · gpt-5-mini · gpt-5-nano
o-series Reasoning	o3-deep-research · o3 · o4-mini
GPT-4 (Legacy)	gpt-4.1 · gpt-4o · gpt-4o-mini · gpt-4-turbo · gpt-4-0613
GPT-3.5 (Legacy)	gpt-3.5-turbo-0125

🔑 Get API key

---

🟠 Anthropic Claude

Series	Models
Claude 4 (Latest)	claude-opus-4-6 · claude-sonnet-4-6 · claude-haiku-4-5-20251001
Claude 3.5 (Legacy)	claude-3-5-sonnet-20241022 · claude-3-5-haiku-20241022
Claude 3 (Legacy)	claude-3-opus-20240229 · claude-3-sonnet-20240229 · claude-3-haiku-20240307
Claude 2 (Legacy)	claude-2.1 · claude-2.0

🔑 Get API key

---

⚪ xAI Grok

Series	Models
Grok 4 (Latest)	grok-4.2 · grok-4.1-fast · grok-code-fast-1
Grok 2 (Legacy)	grok-2-1212 · grok-2-mini

🔑 Get API key

---

🔴 Mistral AI

Series	Models
Frontier (Latest)	mistral-large-2411 · mistral-medium-2508 · magistral-medium-1.2
Legacy	mistral-large-2407 · mistral-small-2409 · open-mixtral-8x22b · open-mixtral-8x7b

🔑 Get API key

---

🔵 Meta Llama (via Together AI)

Series	Models
Llama 4 (Latest)	llama-4-behemoth · llama-4-maverick · llama-4-scout
Llama 3 (Legacy)	llama-3.3-70b-instruct · llama-3.2-90b-vision · llama-3.1-405b · llama-3.1-70b · llama-3.1-8b
Llama 2 (Legacy)	llama-2-70b-chat · llama-2-13b-chat · llama-2-7b-chat

🔑 Get Together AI key

---

🟣 Alibaba Qwen (via Together AI)

Series	Models
Qwen 3 (Latest)	qwen-3.5-plus · qwen-3.5-flash · qwen-3-max-thinking · qwen3-coder-480b
Qwen 2 (Legacy)	qwen2.5-72b-instruct · qwen2.5-coder-32b · qwen2-72b-instruct
Qwen 1.5 (Legacy)	qwen1.5-110b · qwen1.5-72b-chat

🔑 Get Together AI key

---

🚀 Features

Feature	Description
🤖 7 AI Providers	Gemini · GPT · Claude · Grok · Mistral · Llama · Qwen
📂 Multi-format	CSV, Excel (.xlsx/.xls), JSON
💬 Natural language Q&A	Ask anything about your data in plain English
📊 Auto visualizations	AI picks the best chart for your question
🎨 Custom chart builder	Full control over chart type and axes
🔍 Data explorer	Filter, search, and download your data
🧠 AI summary	Executive summary generated by your chosen LLM

---

🎯 How to Use

1. Choose a provider from the sidebar dropdown
2. Select a model from the list
3. Paste your API key → click Connect
4. Upload your data file (CSV, Excel, or JSON)
5. Explore the Dashboard, Chat, Charts, and Raw Data tabs

For Meta Llama and Alibaba Qwen, use a Together AI key instead of a direct provider key.

---

⚙️ Local Setup

# Install dependencies
pip install -r requirements.txt

# Run the app
streamlit run app.py

App opens at http://localhost:8501

---

🏗️ Project Structure

datamind-agent/
├── app.py              # Streamlit UI
├── core_agent.py       # Multi-LLM logic + visualizations
├── requirements.txt    # Dependencies
├── sample_data.csv     # Test dataset
└── README.md           # This file

---

🏗️ Architecture

User (Streamlit UI)
       │
       ▼
  app.py (UI Layer)
       │
       └── core_agent.py
               ├── PROVIDERS{}           → 7 providers, all models, colors
               ├── get_llm()             → LangChain wrapper for any provider
               ├── validate_llm()        → Connection ping test
               ├── load_file()           → CSV / Excel / JSON → DataFrame
               ├── profile_dataframe()   → Statistical profiling
               ├── ask_agent()           → LLM → Answer
               ├── make_plotly_chart()   → Interactive visualizations
               └── ai_recommend_chart()  → LLM picks best chart type

LangChain integrations:
  langchain-google-genai  →  Gemini
  langchain-openai        →  GPT · Grok · Llama · Qwen (OpenAI-compatible)
  langchain-anthropic     →  Claude
  langchain-mistralai     →  Mistral

---

Built with ❤️ using LangChain · Gemini · GPT · Claude · Grok · Mistral · Llama · Qwen · Streamlit