The problem Credix solves

Private credit markets today are both inefficient and vulnerable to discrimination. Loan origination, matching, and servicing rely on slow, manual coordination between intermediaries, creating delays, higher costs, and settlement risk. At the same time, borrowers are often required to share sensitive personal and financial information broadly, enabling discriminatory practices like redlining, where lending decisions are influenced by race, ethnicity, or other non-financial factors. Credix solves both problems by enabling real-time, automated lending on-chain while keeping borrower data private and visible only when necessary, improving efficiency and settlement speed while reducing the ability for lenders to make biased decisions based on identity rather than creditworthiness.

Challenges we ran into

The developer experience with Canton was very unique. Canton is a very different L1 from other blockchains and so building on Canton felt very different from building with Ethereum or other L1s. It was definitely a learning experience writing code in DAML and understanding the unique Canton architecture compared to Ethereum (the diagrams on the docs that the Canton devrels pointed me towards were very helpful). Also, the LLM feature in the docs was very useful (thanks Shreyas for the tip).

Use of AI tools and agents

We used Cursor and Claude Code for coding, and we used ChatGPT and Gemini for planning and ideation.