mnds18/agentic-ts-forecasting-system
0
1---2title: Agentic TS Forecasting System3emoji: 📈4colorFrom: pink5colorTo: indigo6sdk: streamlit7sdk_version: "1.33.0"8app_file: streamlit_app.py9pinned: true10---11 12 13[](https://github.com/mnds18/agentic-ts-forecasting-system/actions/workflows/python-app.yml)14[](LICENSE)15[](https://mnds18-agentic-ts-forecasting-system.streamlit.app)16[](https://huggingface.co/spaces/mnds18/agentic-ts-forecasting-system)17 18# Agentic Time Series Forecasting System19 20This project showcases a modular, multi-agent GenAI system for time series forecasting using a retail sales use case. Designed for enterprise readiness and deployable on local machines, Docker, or cloud platforms.21 22---23 24## 🚀 Key Features25- **🧠 Agent-Oriented Architecture** — Each agent (BA, DS, Reviewer, MLOps, PM, Exec) mimics real roles26- **📈 Model Variety** — Prophet, ARIMA, and XGBoost (optional extensions)27- **🧪 Robust Testing** — Pytest-based unit tests with CI via GitHub Actions28- **📊 Live Dashboard** — Streamlit app with visual charts, error metrics, and execution logs29- **🔁 REST API** — Forecast serving via local Flask30- **📦 Docker Support** — Easily containerized for local/cloud deployment31- **📝 Markdown Reporting** — Agent-generated reports, plans, and presentations32- **🌐 GitHub Pages** — Public-facing docs, walkthroughs, and guides33 34---35 36## 💻 Quick Start37 38### 1. Clone and Set Up39```bash40git clone https://github.com/mnds18/agentic-ts-forecasting-system.git41cd agentic-ts-forecasting-system42pip install -r requirements.txt43```44 45### 2. Run Agentic Workflow46```bash47python agentic_ts_forecasting_system.py48```49 50### 3. View Streamlit Dashboard51```bash52streamlit run streamlit_app.py53```54 55### 4. Query Forecast via REST API56```bash57curl http://localhost:5001/forecast58```59 60---61 62## 🧪 Testing63Run all agent-level tests:64```bash65pytest -v test_agents.py66```67 68---69 70## 📦 Docker Deployment71```bash72docker build -t agentic-ts-app .73docker run -p 5001:5001 agentic-ts-app74```75 76---77 78## 🤗 Hugging Face Space79You can deploy this project to a Hugging Face Space:80- Create a new space under your account81- Set `sdk: streamlit`, point to `streamlit_app.py`82- Add required `requirements.txt` and optionally `Dockerfile`83 84👉 View: [https://huggingface.co/spaces/mnds18/agentic-ts-forecasting-system](https://huggingface.co/spaces/mnds18/agentic-ts-forecasting-system)85 86---87 88## 📘 Documentation89- [WIKI.md](WIKI.md) — System architecture, agent logic, flow diagrams90- [CONTRIBUTING.md](CONTRIBUTING.md) — Dev setup, issue flow, PR standards91- [LICENSE](LICENSE) — MIT License92 93---94 95## 🙌 Credits96Built with ❤️ by Mrig Debsarma and contributors.97 98> Inspired by enterprise-grade AI adoption in forecasting, orchestration, and intelligent automation.99 100---101 102## 📌 Welcome & Community103 104Hi there 👋 — and welcome to one of the most advanced open-source projects in **agentic AI and time series forecasting**!105 106This repo showcases an enterprise-ready, modular GenAI system using a multi-agent workflow, with live dashboards, documentation, Docker packaging, and Hugging Face deployment.107 108### 🔍 What You’ll Find Here109- **Streamlit Dashboard:** 📊 [Live Forecasting App](https://mnds18-agentic-ts-forecasting-system.streamlit.app)110- **Hugging Face Space:** 🤗 [Streamlit app](https://huggingface.co/spaces/mnds18/agentic-ts-forecasting-system)111- **Docs and Guides:** 📘 [WIKI.md](./WIKI.md), [CONTRIBUTING.md](./CONTRIBUTING.md)112- **Agentic Workflow:** 🧠 [agentic_ts_forecasting_system.py](./agentic_ts_forecasting_system.py)113- **API Endpoint:** `GET /forecast` → [localhost:5001/forecast](http://localhost:5001/forecast)114 115### 🧪 How to Contribute116- Fork the repo117- Follow [CONTRIBUTING.md](./CONTRIBUTING.md)118- Run tests with `pytest`119- Open your first PR 🚀120 121### ⭐ Support This Project122- Star 🌟 the repo123- Share it with your network124- Submit issues for bugs, ideas, or improvements125 126Say hi below 👇 or request your first issue to be assigned.127Let’s build the future of autonomous AI workflows together!128