ML Trader
| Pillar | Status |
|---|---|
| Build | Active |
Description
ML Trader is our trading system based on machine learning models used to predict the future prices of financial assets, with the goal of generating capital gains that fund Agenticus’ social projects. Currently, with the returns obtained, we are covering our operating expenses and directly financing the project Agenticus Scholarships.
Arquitecture of Our System
Our System is built on GCP and is made of theses components:
1. A virtual machine that automaticly gathers daily stock market data into Buckets.
2. A datamart built on Big Query.
3. ML models trained in Work Bench and registered in the Model Registry.
We are continuously developing and improving our technology; and of course, in line with our open source mindset, we are sharing the code in the GitHub repository.
Why ML Trader Exists?
During the first year of Agenticus, the costs and expenses were covered by the personal income of its founder Fernando Rodriguez. This approach worked for implementing the first projects of Agenticus, however it was not sustainable when trying to work on bigger and more impactful projects. After a few-months pause in which we discussed and evaluated how to make Agenticus economically self-sustainable, we finally decided to use data and AI to generate income. That is how ML Trader was born.
Volunteers:
- Fernando Rodriguez - developer
- Christian Ventura - developer
- Wendy Rivera - financial analyst