🤖 Autonomous Data Analyst

An AI-powered Multi-Agent Analytics Platform that automates the complete analytics lifecycle—from live data ingestion and statistical hypothesis generation to AI-powered executive reporting and historical market intelligence.

---

🚀 Overview

Autonomous Data Analyst is an AI-powered analytics platform that automates the complete business intelligence workflow.

Instead of manually exploring datasets, the platform employs multiple AI agents that collaborate to:

• Collect live cryptocurrency market data
• Profile datasets automatically
• Generate statistical business hypotheses
• Execute statistical analysis
• Validate analytical results
• Generate executive business reports
• Store historical market intelligence
• Compare historical trends
• Present everything in an interactive dashboard

The project combines modern Data Engineering, Business Intelligence, Statistics, Generative AI, and Agentic AI into a unified analytics platform.

---

🎯 Project Impact

• ⚡ Automated an end-to-end analytics workflow, reducing manual exploratory analysis and hypothesis generation by approximately 80%.
• 🤖 Eliminated repetitive manual reporting by automatically generating executive summaries and business recommendations using AI agents.
• 📊 Reduced the time required to transform raw cryptocurrency market data into actionable business insights from hours to minutes.
• 🧠 Enabled historical market intelligence by storing and comparing previous analytical snapshots through a dedicated Memory Agent.

---

✨ Features

📡 Live Data Collection

• CoinGecko API Integration
• Live Cryptocurrency Market Data
• Automated ETL Pipeline
• Bronze → Silver Architecture

---

📊 Data Engineering

• Automated Data Cleaning
• Feature Engineering
• SQLite Data Warehouse
• Historical Market Storage

---

🤖 Multi-Agent AI Workflow

The project uses a LangGraph-based Agentic AI workflow consisting of six specialized agents.

📊 Profiler Agent

• Dataset Profiling
• Missing Value Detection
• Duplicate Detection
• Correlation Analysis
• Outlier Detection
• Summary Statistics

---

💡 Hypothesis Agent

Powered by Google Gemini.

Automatically generates statistically testable business hypotheses based on the dataset profile.

Examples:

• Price vs Market Cap
• Volume vs Market Cap
• Rank vs Price
• Correlation Analysis
• Regression Problems

---

📈 Analyst Agent

Executes statistical tests including:

• Pearson Correlation
• Spearman Correlation
• Linear Regression
• T-Test
• ANOVA

---

✅ Critic Agent

Reviews every analytical result.

Checks:

• Statistical correctness
• Business relevance
• Confidence level
• Recommendation quality

Only approved insights move forward.

---

📝 Narrator Agent

Transforms statistical outputs into executive-level business reports.

Generates:

• Executive Summary
• Key Findings
• Business Recommendations
• Final Conclusion

---

🧠 Memory Agent

Stores every market snapshot inside SQLite.

Tracks:

• Market Capitalization
• Trading Volume
• Bullish/Bearish Trend
• Historical Executive Reports

Allows historical comparison across pipeline executions.

---

🏗 Project Architecture

                 CoinGecko API
                       │
                       ▼
              Bronze Data Layer
                       │
                       ▼
              Silver Data Layer
                       │
                       ▼
             SQLite Data Warehouse
                       │
                       ▼
               LangGraph Workflow
                       │
      ┌───────────┬─────────────┬─────────────┐
      ▼           ▼             ▼             ▼
 Profiler   Hypothesis     Analyst       Critic
      │
      ▼
  Narrator
      │
      ▼
   Memory Agent
      │
      ▼
 Streamlit Dashboard

---

📷 Application Screenshots

🏠 Application Overview

---

📊 Cryptocurrency Dashboard

Live cryptocurrency KPIs, AI executive summary, key findings, recommendations, and market insights.

---

🤖 AI Analysis Report

Displays the complete LangGraph workflow including Profiler, Hypothesis, Analyst, Critic, Narrator, and Memory Agents.

---

🧠 Memory Agent

Compares the latest market snapshot with previous pipeline executions and tracks historical market intelligence.

---

📜 Market History

Historical market trends, pipeline executions, sentiment evolution, and executive summaries across runs.

---

ℹ️ About the Project

Project architecture, AI workflow, technology stack, features, and overall system design.

📂 Project Structure

autonomous-data-analyst/

├── agents/
├── config/
├── data/
│   ├── bronze/
│   └── silver/
├── database/
├── images/
├── ingestion/
├── memory/
├── models/
├── orchestrator/
├── pages/
├── reports/
├── ui/
├── utils/
├── warehouse/
│
├── app.py
├── requirements.txt
├── README.md
└── .env.example

---

📊 Dashboard Modules

📊 Dashboard

• Live KPIs
• Market Capitalization
• Trading Volume
• Executive Summary
• AI Recommendations

---

🤖 AI Analysis

Displays outputs from every AI Agent:

• Profiler
• Hypothesis
• Analyst
• Critic
• Narrator
• Memory

---

🧠 Memory

Historical comparison between the latest and previous market snapshots.

---

📜 History

Tracks historical pipeline executions.

---

ℹ About

Complete architecture, technology stack, workflow and project details.

---

🛠 Technology Stack

Layer	Technologies
Programming	Python
Data Engineering	Pandas, NumPy
Statistical Analysis	SciPy, Statsmodels
AI Framework	LangGraph
LLM	Google Gemini
Database	SQLite
Visualization	Plotly
Frontend	Streamlit
API	CoinGecko API

---

⚙ Installation

Clone the repository

git clone https://github.com/swapitsneil/autonomous_data_analyst.git

Move into the project

cd autonomous_data_analyst

Install dependencies

pip install -r requirements.txt

Create a .env file

GEMINI_API_KEY=YOUR_GEMINI_API_KEY
COINGECKO_API=YOUR_COINGECKO_API_KEY

Run the ETL Pipeline

python -m ingestion.api_loader

Launch the Streamlit application

streamlit run app.py

---

📈 Example Workflow

Fetch Live Market Data

↓

Clean & Transform Dataset

↓

Store in SQLite

↓

Profiler Agent

↓

Hypothesis Agent

↓

Analyst Agent

↓

Critic Agent

↓

Narrator Agent

↓

Memory Agent

↓

Interactive Dashboard

---

🔮 Future Roadmap

• PDF Executive Report Export
• Multi-Dataset Support (CSV, Excel, SQL)
• Smarter Temporal Memory
• Visualization Agent
• Auto Dashboard Generation
• Scheduled Pipeline Execution
• Docker Deployment
• Cloud Database Support
• User Authentication

---

👨‍💻 Developer

Swapnil Nicolson Dadel

Generative AI Data Analyst

• 💼 LinkedIn: https://www.linkedin.com/in/swapnildadel/
• 🐙 GitHub: https://github.com/swapitsneil

---

⭐ Support

If you found this project useful, consider giving it a ⭐ on GitHub.