Under the Hood

Discover the technology and infrastructure powering our strategies

Features at a Glance

An integrated system that takes models from idea to live deployment with minimal manual work

Automated ML Pipeline: In-house framework that uses simple JSON configs to handle full model lifecycles: dataset downloads, parsing, preprocessing, training from linear regressions to neural networks, evaluation, and seamless inference.
Intelligent Task Scheduler: Python scheduler, automates fetching latest market metrics, calculating strategies, and triggering GCP workers to generate predictions
Dynamic Dashboards & Web: Sleek web & dashboards powered by modern web technologies like Sveltekit
Scalable Cloud Inference: Python-based machine learning and internal tooling run on Google Cloud Platform for efficient model inference

Supabase, Python, and SvelteKit are our essential building blocks that power our infrastructure

Python

Core backend for machine learning pipelines, data processing, API development, and automated prediction strategies
SvelteKit

High-performance frontend framework powering reactive websites, dynamic dashboards, and insight portals
Supabase

Real-time database, authentication, and storage for real-time dashboards and secure user management across all projects

Rigorous 3-phase journey from research to production strategies

Phase 1

Research & Backtesting

Train models on historical data with extensive diagnostics and risk analysis

Phase 2

Live Shadow Testing

Deploy live predictions to internal dashboards without staking capital

Phase 3

Production Strategies

Promote proven models to continuously contribute to live strategies

Our development pipeline for continuous improvement