ML in Blockchain / on-chain credit scoring

🌀

ML in Blockchain / on-chain credit scoring

🤔

Our group is solving the problem of unequal access to credit. Our North Star is an on-chain self-governed reputation scoring mechanism.

💡

We’re looking for an Analytics Team Lead with prior experience in Data Science, ML Engineering; credit scoring & credit analysis. However, no expertise in blockchain dev is necessary.

What we offer

Learn the blockchain tech and dive deeper into mechanisms of lending and borrowing smart contracts - Web3.py, Brownie

Acquire skill in accessing and downloading blockchain lending data from Aave and Compound smart contracts - Graph CLI, geth/erigon

Build scalable data engineering pipeline (python stack, Spark)

Display created features using, web dev, (FastAPI interface, React, PostgreSQL, docker)

See how to apply your machine learning skills to improve the capital efficiency of on-chain lending/borrowing transactions (LightGBM, SHAP, time-series + CNN, causal modeling)

Goal

In the next 8 weeks, we’re aiming to build a tool that analyzes the history of blockchain user transactions and suggests optimal collateral lower than the market, keeping the same amount of risks.

Core Team

notion image

Pasha Datsiuk | Group Lead Revenue Ops at Purefacts

notion image

Daniel Kur | Data Team Lead Applied Research Scientist at ServiceNow

notion image

Edward Johnson | Core Team Founder at Ikinique

notion image

Solito Reyes III | Core Team Data Scientist at Vegito

notion image

Chris Li | Core Team Technology Analyst

notion image

Shubhanker Banerjee | Core Team Researcher at Adapt Center

Contributors

notion image

notion image

Pablo Benavides

notion image

Yağmur Yılmaztürk

notion image

notion image

notion image

notion image

notion image

notion image

❗

Timeline

Our wg is divided into multiple teams, thus most of work is done in parallel

#	Major Milestones	Expected time to finish
1	Onboarding	2 weeks
2	Data preparation	done
3	Featurization	2 weeks
4	Get to a baseline model	3 weeks

Our Published Work

https://ai.science/l/ac6d4904-b9f1-4630-8d5b-d07fc020d918

On-chain lending data aggregation and analysis

This article focuses on the challenges and solutions we faced when building a robust on-chain reputation scoring mechanism within the collateralized DeFi segment. At the current stage, we are building a dataset, based on Aave that unifies health scores across lending protocols by aggregating and analyzing on-chain lenders' data.

https://medium.com/incyd/on-chain-lending-data-aggregation-and-analysis-33fe6f469b17

Web site created using create-react-app

http://incyd.io/

Why join?

Experience 0 to 1 from within a startup

Learn about the forefront blockchain tech

See how your ml/blockchain skills can be applied in one of the most fast-growing areas

[Recruitment on hold. But you can join the group’s public Slack channel here for future opportunities.]