EcoCode AI

results
simulator
gemini ai
dashboard
repo scan

Inspiration

As AI systems and cloud workloads continue to grow, inefficient software increasingly contributes to energy consumption and digital carbon emissions. We built EcoCode AI to help developers understand the environmental impact of their code and make sustainable software engineering measurable, practical, and accessible.

What it does

EcoCode AI is an AI-powered sustainable software engineering platform that analyzes Python code and GitHub repositories to detect inefficient execution patterns, estimate environmental impact, and generate intelligent optimization suggestions.

The platform provides:

Estimated CO₂ emission tracking
Runtime and memory analysis
Code complexity analysis
Green Score generation
AI-powered optimization suggestions
Before-vs-after benchmarking
Repository-wide sustainability scanning
Secure sandboxed code execution

How we built it

We built EcoCode AI using Python and Streamlit for the frontend and interactive dashboard experience.

Our analysis pipeline combines:

AST parsing for static code analysis
Radon for complexity metrics
psutil and tracemalloc for runtime monitoring
Plotly for interactive benchmarking charts
Secure subprocess sandboxing for isolated code execution

We integrated the Google Gemini API to power intelligent sustainability analysis and code optimization suggestions. Gemini analyzes inefficient code patterns, explains their environmental impact in plain language, and generates optimized versions focused on reducing runtime, memory usage, and computational waste.

Challenges we ran into

One of the biggest challenges was securely executing untrusted code while accurately collecting runtime and sustainability metrics.

We also faced challenges in:

estimating emissions for short-running scripts
creating meaningful optimization comparisons
ensuring AI-generated optimizations remained transparent and understandable
balancing optimization quality with safe execution

Accomplishments that we're proud of

We are proud of building a fully working AI-powered sustainability platform that transforms invisible software inefficiencies into measurable insights.

Some accomplishments include:

AI-powered optimization workflow using Gemini API
Real-time sustainability benchmarking
Interactive before-vs-after simulator
Repository-wide sustainability analysis
Secure sandboxed execution engine
Transparent Green Score system
Interactive dashboards and PDF sustainability reports

What we learned

We learned that sustainable software engineering is closely connected to efficient software engineering. Reducing unnecessary computation not only improves performance and scalability, but also lowers energy consumption and digital carbon emissions.

We also gained experience with: