DecTell AI - Decision Intelligence Platform

An AI-powered Business Analytics & Decision Intelligence platform built with Python and Streamlit. Upload any dataset, get LLM-generated insights, train ML models, simulate decisions, and chat with your data all in one place.

📸 Screenshots

Features

Module	Description
Smart Data Upload	AI-driven cleaning - imputes, encodes, caps outliers only when needed
Automated EDA	Distribution charts, correlation heatmaps, LLM-generated narrative insights
AI Analyze	Business mode (plain language goals) + Technical mode (full ML control)
Scenario Simulation	What-if sliders and dropdowns for any dataset type including categoricals
Causal Impact	Difference-in-Differences with bootstrapped 95% confidence intervals
Chat with Data	Groq LLM conversational interface with full session memory
Report Generator	AI-written Business and Technical reports exportable as PDF or text
Analytics Dashboard	Live KPIs, model metrics, feature importances
Developer Page	Professional profile of the developer

---

Tech Stack

• Language: Python 3.12
• Framework: Streamlit
• AI Engine: Groq API with llama-3.3-70b-versatile
• ML: Scikit-learn, XGBoost
• Data: Pandas, NumPy
• Visualisation: Plotly
• PDF Export: ReportLab
• Integration: Google Sheets API

---

Local Setup

1. Clone the repository

git clone https://github.com/your-username/dectell-ai.git
cd dectell-ai

2. Install dependencies

pip install -r requirements.txt

3. Add your secrets

Create .streamlit/secrets.toml locally - this file is gitignored and never pushed:

GROQ_API_KEY       = "gsk_your_groq_api_key_here"
SHEETS_WEBHOOK_URL = "https://script.google.com/macros/s/YOUR_SCRIPT_ID/exec"

Get your free Groq API key at console.groq.com

4. Run the app

streamlit run app.py

---

Secrets Reference

Key	Where to get it	Required
GROQ_API_KEY	console.groq.com - free tier available	Yes - for all AI features
SHEETS_WEBHOOK_URL	Google Apps Script deployment URL from your spreadsheet	No - only for contact and feedback forms

---

Deploy on Streamlit Cloud

1. Push this repo to GitHub
2. Go to share.streamlit.io
3. Click New app and select your repo
4. Set main file as app.py
5. Under Advanced settings → Secrets paste your secrets in TOML format
6. Click Deploy

---

Partner

Data & Disco Dreams Studio Data Consulting & Design for Small Businesses | Atlanta, United States

---

Developer

Vivek Singh MCA - Big Data & Analytics | Jaypee Institute of Information Technology, Noida

• LinkedIn: vivek-singh-linkdin
• GitHub: vivek081202
• Email: vivekkrsingh082003@gmail.com

---

Licence

MIT - Free to use, modify, and distribute.

---

DecTell AI - Built with research, analytics, and intelligence.