# Risk management

### Underwriting

We underwrite the user based on:\
\- **Onchain data**: wallet assets, transaction history, DeFi activity; \
\- **Offchain data**: bank account balance, assets & liabilities, cashflow;\
\- **Exotic data sources** like criminal history, tax reports, and employment status, that we get with **zkTLS** or **Open Banking**;\
\
Using this mix allows us to get more **accurate results** than most web2 companies, as well as **serve a global audience**. 

We design **tailored algorithms** (decision tree, weights, and thresholds) **for each underwriting case**, based on the specifics of our partner integration: location, userbase, and payment target.

### Servicing & default management

We currently allow user to **repay their debt manually** or **inside our partner app** (through API). For EVM networks, we leverage our partner's infrastructure to enable **automatic recurring payments** (see [Otim Docs](https://docs.otim.com/docs/use-cases#recurring-transfers)). We plan to use more onchain protocols to optimize repayment UX in the future. 

We employ several **offline debt collection agencies** to service our debt if needed. Special agreements with our partners allow us to **restrict app or asset access** in case of delinquency.