Presentation: https://docs.google.com/presentation/d/12cXucCG3TLk3vAvNZVrFgRP5LdEcB7-S/edit?usp=sharing&ouid=113904595829423638187&rtpof=true&sd=true

Inspiration

Global events such as the Russia-Ukraine war and natural disasters like the LA wildfires got us thinking about how companies might be affected. In particular, how do companies anticipate and respond to supply chain disruptions? In addition, how can companies improve their supply chain to reduce their carbon footprint and source materials more sustainably?

What it does

ChainReact utilizes Aryn to analyze a diverse array of data, including news and current events, weather reports and forecasts, trade and sanction reports, information on natural disasters, etc. This data is processed and stored in a Singlestore vector database. Companies can log in to our web application and submit company details and details about their supply chain such as supplier information, inventory, logistics and transportation routes, etc. ChainReact generates a query for a LLM (Anthropic) based on this information and applies a similarity search in the database, identifying possible disruptions to the company's supply chain and providing actionable steps to mitigate the impact of these disruptions, as well as ways to reduce carbon footprint and source materials more sustainably.

How we built it

We used Aryn to parse pdf documents from the web, Singlestore to store the vector database, and Terraform to host the EC2 interfaces.

What we learned

We learned a lot about supply chains in general and what challenges companies face, and how they respond. We learned how to use Aryn and Singlestore. Also, 24 hours is not a very long time.

What's next for ChainReact

  • Improve the app's UI and specifically, include map functionality so things like natural disasters possibly affecting a company's supply chain are displayed.
    • Implement a web-scraper that pulls articles and documents from various internet sources as they are published, in real-time.
    • Give each company a "resilience score" based on the reliability of their supply chain, so that once many companies have ChainReact accounts, more accurate supply chain analyses can be made.

Built With

Share this project:

Updates