AutoGS

AutoGuided Onboarding WebApp for Carbon Footprint Reduction (v3.0)

Powered by AutoGS AI & AWS SageMaker

---

This solution includes:

1. Wireframe Design
2. Prototype
3. Mockup Design
4. SageMaker AI Model Update
5. Development Requirements
6. Getting Started
7. Contribution Guidelines
8. Code of Conduct
9. Contact
10. License
11. Resources

---

SageMaker AI Model Update

Pipeline Overview

• Training: Uses SmallSat earth observation data, processed via imagery_processing.ipynb
• Model Registry: Tracks and versions trained models with full audit logs
• Deployment: Real-time inference through SageMaker Endpoints connected to the dashboard
• AutoML Option: Supports H2O AutoML for automated model selection

Key Files

• ml/carbon_model_sagemaker_pipeline.ipynb: Defines the SageMaker ML pipeline
• ml/deploy_endpoint.py: Deploys models to SageMaker Endpoints
• ml/inference_handler.py: Handles API requests from the frontend

AWS Services Used

• Amazon SageMaker
• AWS Ground Station
• Amazon S3
• AWS Lambda
• Amazon CloudWatch

Benefits of Migration

• Managed scalability
• Built-in versioning
• Secure infrastructure
• Reduced operational overhead

---

Development Requirements

• Python 3.9+
• AWS CLI configured (IAM permissions required)
• AWS SDKs: boto3, sagemaker
• Python packages:
  • numpy, pandas, scikit-learn, matplotlib, transformers
  • torch, tensorflow, gradio, rasterio, cv2
  • sagemaker, boto3
• Optional: Docker (for local SageMaker testing)
• Git client

---

Getting Started

1. Clone the repository:

git clone https://github.com/aimtyaem/autogs.git
cd autogs

2. Set up Python environment:



python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt

3. Configure AWS credentials:

aws configure

4. Run SageMaker pipeline notebook:

Open ml/carbon_model_sagemaker_pipeline.ipynb and execute cells.

Deploy endpoint with deploy_endpoint.py.



5. Launch frontend and connect to live AI models.

---

Contribution Guidelines

---

Code of Conduct

---

Contact

Ahmed Ibrahim Metawee

---

License

Licensed under the MIT License.