Inspiration

It is essential for potential projects to qualify for regional ESG requirements. Without meeting ESG requirements, projects will face roadblocks to getting funding, thus delaying deployment of impactful renwable energy technologies.

To meet these requirements, project contractors consult with various consultants and engage in manual dilligence. The dilligence process takes extended amounts of time and is very expensive (1-3 years before building starts!).

To accelerate this process, we created an all-in-one ESG consultant that gives an overall score based off general project information, and provides a more detailed analysis on Environmental, Social, and Governance implications for the project.

What it does

Select your City, State. Enter your project type, energy output, and cost. Generates an ESG score and report estimate from consulting with provided ESG resources.

Generates a dashboard with renewable energy projects. On the dashboard, you can generate, and chat with an E, S, and G agent that provides tailored information for your project based on location. The E agent is backed with biodiversity and geolocated information for the project. The S agent is connected to census information and marketing tips for various demographics. The G is connected to a comprehensive document on the permits and governance requirements for specific locations.

How we built it

ESG -> OpenAI LLM connected to FEMA, ESG survey, and ESG survey prediction E agent -> OpenAI LLM connected to biodiversity database for various areas. S agent -> OpenAI LLM connected to census information and trained on specific marketing campaigns. G agent -> RAG Agent connected to a comprehensive database for governance permits. Streamlit for backend Docker for frontend Development: Cursor.AI IDE Database: Supabase Web Scraping: crawl4ai AI Integration: OpenAI

Challenges we ran into

Data is not consistent for various states, cities, and locations. Websites are not scrapeable in the same way. LLMs hallucinate unless very finely tuned.

Accomplishments that we're proud of

Multiple agents communicating with eachother. Accurate ESG scores based on locations. RAG agent for specificly catered information for specific cities and locations.

What we learned

How to build agents rapidly and create value adding interfaces.

What's next for Vamm.eco

Adding more information on the money saved, and the economic impact for the areas for the projects. As well as a streamlined way to provide a social media marketing campaign.

Built With

  • crawl4ai
  • docker
  • openai
  • streamlit
  • supabase
Share this project:

Updates