AWS Cloud Sustainability Optimiser

Optimise AWS code/IaC for sustainability, reduce carbon footprint. This tool will analyse, measure, modify, and provide recommendations. It's time to build responsibly to save the world for future!

Comment

Inspiration

I do remember when I started coding my main focus was on clean coding practices and how to make it performant, encapsulated and so on. That has become a habit now if I have learn anything from it :D

These days I am trying to focus on Sustainability aspects of the code I write or the solutions I design. As a proof of concept, I created the “AWS Sustainability Optimiser” application. This app will easily help you to evaluate the carbon impact of your code, your solution architecture design on the environment according to “Best practices for sustainability in AWS”.

For instance did you know that per DynamoDB call carbon cost is 0.00005kg CO2e and 0.00002kg CO2e per SQS call! Me neither. But the app helped me to find it out and corrected my code so that I publish less SQS messages to reduce the generated carbon footprint.

What it does

The app will analyse your current cloudformation solution, or your code. It will provide suggestions on how to improve the code to reduce the carbon footprint impact. It Estimates the Carbon footprint Impact of your original and optimised solution. You can fully compare how a better solution can benefit our environment.

How we built it

I started with the idea of who are the personas for this app, and how they might want to use it. The personas for this tool could be developers, Solution Architects and AWS Cloud administrators. You can provide your code snippets, or the IaC for your workload resources (Cloudformation, CDK,..). Based on that I started improving the prompt to create more accurate responses.

Challenges we ran into

My initial intention was to include the Infrastructure Diagram as well, but that was too challenging to get the model to read the image. My next challenge was to ask model to generate Analytics on how the original solution impacted whereas the recommended solution. it's challenging to create accurate diagrams and charts.

Accomplishments that we're proud of

The more I play with Amazon Bedrock and Party Rock I get to know more about Prompts, prompt templating and I believe that's been a great achievement for me. the whole purpose of this contribution was to learn more and try something new. I am proud I did it.

What we learned

Great opportunity to know about importance of the prompt engineering. LLM models

What's next for AWS Cloud Sustainability Optimiser

Built With

  • partyrock
Share this project:

Updates