Crux Climate Data Insights Platform

Inspiration

Crux Climate is pioneering the ecosystem for transferable tax credits in clean energy development. However, as a new market, they face challenges due to limited data. Our team was inspired to create tools that would help Crux overcome this hurdle and make more informed decisions in this emerging field.

What it does

Our project, Carbon Connect, is a comprehensive suite of tools designed to generate, analyze, and visualize data for the transferable tax credit market. The platform includes:

Synthetic Data Generation: Uses machine learning and LLMs to create realistic synthetic data for anomaly detection and market simulation. It then allows for ad-hoc distribution matching.
Data Visualization Dashboard: Provides interactive visualizations of buyers, bids, prospective buyers, and their aggregated data.
Key Metric Analysis: Offers in-depth analysis and visualization of critical market metrics.
AI-Powered Clustering and Matching: Utilizes embeddings to cluster buyers and credits, then matches them based on bid data.
Suggested buyers for credits: Scraped 1000 largest US profitable public companies for their revenue, tax liability and industry.

How we built it

We developed our platform using a combination of:

Machine Learning algorithms for synthetic data generation
Large Language Models (Claude 3) for advanced data synthesis and analysis
OpenAI embedding model for clustering credits, and buyers.
Data visualization libraries for creating interactive dashboards
Embedding techniques and clustering algorithms for buyer and credit analysis
Matching algorithms to pair buyers with appropriate credits

Challenges we ran into

The main challenge was working with limited real-world data in this new market. We had to ensure our synthetic data generation was realistic and useful for Crux's needs. Additionally, creating meaningful clusters and matches with limited historical data required innovative approaches.

What we learned

We gained deep insights into the emerging transferable tax credit market and the unique challenges it presents. We also learned about advanced techniques in synthetic data generation, clustering, and matching algorithms in the context of limited data scenarios.

What's next for Crux Climate Data Insights Platform

Moving forward, we plan to:

Refine our synthetic data generation to incorporate more complex market dynamics
Enhance the matching algorithm to consider more nuanced factors in buyer-credit pairing
Develop predictive models to forecast market trends as more real data becomes available
Work closely with Crux to integrate our platform into their existing systems for seamless adoption