🔗 Live App: data-analyst-ai-agent.streamlit.app

---

📊 Data Analyst Agent 🤖

Unlock the Power of Data with AI — Instantly Analyze, Visualize, and Discover Insights!

🔗 Live App: data-analyst-ai-agent.streamlit.app

---

🚀 Project Overview

Data Analyst Agent is a next-generation AI-powered analytics platform built with Streamlit. It empowers users—whether beginners, business professionals, or data scientists—to upload their data, clean it, analyze it, visualize it, and even build machine learning models instantly.

The app integrates modern UI/UX design with cutting-edge AI models, providing both automation and interactivity in a seamless way. Users can chat with their data, generate SQL queries, build dashboards, and even perform AutoML—all in one place.

This is not just a tool—it’s your virtual AI Data Analyst.

---

✨ Key Features

🎨 1. Modern, Animated UI/UX

• Dark neon-inspired design with glassmorphism.
• Animated gradients and interactive transitions.
• Sidebar with profile, settings, and navigation.
• Responsive and elegant design for all devices.

---

📂 2. Sidebar Features

• App logo + user profile section.
• Quick navigation across tabs (Upload, Insights, SQL, Visualization, ML, etc.).
• Theme switcher (light/dark/glass).
• Recent file history for quick access.
• API key management (.env or manual entry).
• App settings (language, font size, preferences).
• Social links (GitHub, Twitter, LinkedIn).
• Support, feedback, and developer credit section.

---

📑 3. Data Upload & Preprocessing

• Upload CSV/Excel files with drag-and-drop.
• Automatic type detection, date parsing, and missing value handling.
• Preview dataset instantly.

---

🧹 4. Data Cleaning & Transformation

• Fill, drop, or filter missing values.
• Convert/transform data types interactively.
• Apply filters and transformations directly from UI.

---

📊 5. Automated Insights & Anomaly Detection

• AI-powered anomaly detection.
• Summary statistics at your fingertips.
• OpenAI-powered insights for business storytelling with data.

---

💬 6. Conversational Data Exploration

• Chat with your data like a human analyst.
• Multi-turn memory for context-aware conversations.
• Export chat history, SQL queries, and responses.

---

🧑‍💻 7. SQL Analysis

• AI-generated SQL queries from plain English.
• Query correction with error explanations.
• Visual query builder for no-code SQL.
• Version control and collaborative query sharing.

---

📈 8. Data Visualization Studio

• Automatic chart recommendations.
• Drag-and-drop chart builder.
• Build complete dashboards with filtering & drill-down.
• Export charts as images, data, or dashboards.
• AI-powered annotations & insights.

---

📊 9. Power BI Analyst

• Data modeling & KPI dashboards.
• Auto-generate DAX and Power Query scripts.
• Data source connectors and refresh scheduling.
• Export reports for Power BI.

---

⚡ 10. Advanced BI Tools

• Time intelligence (period over period analysis).
• Drill-through navigation.
• Custom visual marketplace (e.g., radar charts).
• Row-level security simulation.
• Publish & share dashboards securely.

---

🤖 11. Machine Learning Analyst

• AutoML model builder (classification & regression).
• Feature engineering & selection tools.
• Explainability via SHAP values.
• Prediction playground for testing models.
• Export models for reuse.

---

🛠️ Tech Stack & Integrations

• Frontend/UI: Streamlit (dark neon theme with animations).
• Data Processing: Pandas, NumPy, DuckDB.
• LLM Features: OpenAI API.
• Visualization: Plotly, PyDeck.
• Machine Learning: Scikit-learn, SHAP, Joblib.
• Environment Management: .env + Streamlit secrets.

---

📌 Why This Project?

Data analysis is often time-consuming and technical. With Data Analyst Agent, you get the power of a data analyst, BI engineer, and ML engineer all in one friendly AI-powered tool. Whether you’re analyzing sales data, visualizing trends, or building predictive models—this app does it all, instantly.

---

graph TD

A[Start: Launch App] --> B[Upload Data (CSV/Excel)] B --> C{File Uploaded?} C -->|No| B C -->|Yes| D[Preprocessing: type detection, missing values, date parsing]

D --> E[Data Cleaning & Transformation] E --> F[Automated Insights & Anomaly Detection]

F --> G{User Selects Next Action}

G -->|Conversational Q&A| H[Chat with Data] H --> H1[Export Chat + SQL + Results]

G -->|SQL Analysis| I[AI SQL Generator + Visual Query Builder] I --> I1[Export SQL Queries + Tables]

G -->|Visualization Studio| J[Auto Chart Suggestions + Dashboard Builder] J --> J1[Export Charts/Dashboards]

G -->|Power BI Analyst| K[KPI Metrics + DAX + Connectors] K --> K1[Export for Power BI]

G -->|Machine Learning Analyst| L[AutoML Model Builder + Predictions] L --> L1[Export Models + Predictions]

H1 --> Z[End: Download Reports] I1 --> Z J1 --> Z K1 --> Z L1 --> Z

---

🧩 App Workflow (LangGraph + Decision Tree)

Below is the end-to-end workflow of how the app works, represented as a graph of nodes (states) and edges (decisions/actions):

flowchart TD

%% START
A[🚀 Start App\nUser Opens Streamlit App] --> B[📂 Upload Data\n(CSV/Excel)]

%% DATA UPLOAD
B --> C{✅ File Uploaded?}
C -->|No| B
C -->|Yes| D[🔍 Preprocessing\nType inference, missing values, date parsing]

%% CLEANING & TRANSFORMATION
D --> E[🧹 Data Cleaning\nFilter, fill, drop, type change]
E --> F[📊 Automated Insights\nStats + Anomaly Detection + LLM insights]

%% BRANCH OPTIONS
F --> G{📌 Next Step?}

%% Conversational Exploration
G -->|💬 Chat with Data| H[🤖 Conversational Q&A\nMulti-turn memory, export chat]

%% SQL
G -->|🧑‍💻 SQL Analysis| I[🗃️ SQL Query Generator\nAI query + visual builder + versioning]

%% Visualization
G -->|📈 Visualization| J[🎨 Visualization Studio\nAuto charts + dashboard builder]

%% Power BI
G -->|📊 Power BI Analyst| K[⚡ Power BI Tools\nDAX, metrics, connector hub]

%% Machine Learning
G -->|🤖 ML Analyst| L[🧠 AutoML Model Builder\nTrain, explain, predict, export]

%% Endpoints
H --> Z[✅ Download Results]
I --> Z
J --> Z
K --> Z
L --> Z

%% END
Z --> X[🏁 End Session / Export Data & Reports]

Loading

---

🛠 How It Works (Step by Step)

1. 🚀 Launch App → User opens the Streamlit app (live demo).

2. 📂 Upload Data → CSV/Excel file uploaded.

3. 🔍 Preprocessing → Automatic type detection, missing value handling, date parsing.

4. 🧹 Cleaning → User interacts to clean/transform data.

5. 📊 Insights → LLM + stats engine provides anomalies & insights.

6. Decision Node → User chooses next action:

   • 💬 Chat → Natural language exploration.
   • 🗃️ SQL → AI-powered SQL queries & visual builder.
   • 🎨 Visualization → Build charts/dashboards.
   • ⚡ Power BI → Advanced BI & DAX scripts.
   • 🧠 ML Analyst → Train AutoML models.

7. ✅ Export → Download insights, dashboards, queries, models.

8. 🏁 End → Session complete, reports ready to share.

---

---

👨‍💻 Developer Notes

Abhi Yadav This project is actively evolving. Future updates will bring:

• More visualization templates.
• Deeper integrations with BI tools (Tableau, Power BI cloud).
• Natural language → dashboard automation.
• Real-time data streaming support.

---

🌐 Live Demo

👉 Try it here: data-analyst-ai-agent.streamlit.app

---

4. Optionally, export the diagram from https://mermaid.live as a PNG/SVG and upload it to your repo /assets/ folder.

---

❤️ Made with Passion by Abhishek Yadav & Open-Source Contributors! 🚀✨

© LICENSE

This project is licensed under MIT.

---

Thanks for checking out my GitHub Profile!

💌 Sponser

👨‍💻 Developer Information

Created by: Abhishek Kumar
📧 Email: abhiydv23096@gmail.com
🔗 LinkedIn: Abhishek Kumar
🐙 GitHub Profile: @abhishekkumar62000

Don't forget to give A star to this repository ⭐

👍🏻 All Set! 💌

---