Technical Architecture

Overview:

• Frontend: Web-based interface built with React and TypeScript.

  • Includes an IDE-like code editor for advanced users.

  • Chat interface powered by LLM for natural language strategy creation.

  • Visualization components for backtest/live results (charts, graphs).

  • UI elements: sliders, dropdowns, toggles, tables, and searchable marketplace listings.

• Backend: Python-based microservices.

  • Handles strategy parsing, backtesting (using Freqtrade), and live trading.

  • RESTful or GraphQL API layer for authentication, strategy management, and marketplace operations.

  • Token-based authentication (e.g., JWT).

• Containerization & Isolation:

  • All microservices run in Docker containers, orchestrated via Kubernetes.

  • User sessions/strategies isolated in containers to prevent data leaks.

  • Follows Docker best practices (read-only file systems, security profiles).

2. Infrastructure (Hosted on GCP)

• Kubernetes Cluster:

  • Autoscaling handles spikes in load (e.g., during backtests).

  • Load balancers distribute traffic across pods.

  • Secrets managed via Kubernetes Secrets.

• Networking & Security:

  • VPC with private subnets for databases and compute clusters.

  • Public subnets for UI/API gateways.

  • All traffic encrypted with TLS/SSL.

  • Role-based access control (RBAC) for Kubernetes resources.

  • API rate limiting and OAuth 2.0/JWT-based access control.

• Storage & Logging:

  • Historical price data stored in BigQuery or Cloud Storage.

  • Backtest results (CSV, JSON, performance charts) stored in Cloud Storage.

  • Logs and metrics sent to Cloud Logging and Monitoring.

  • Persistent volumes for data, logs, and backtest results.

• Databases:

  • Managed databases (Cloud SQL or Firestore) store user data, strategies, and marketplace metadata.

• Failover & Redundancy:

  • Geo-redundant hosting for high availability.

  • Automated backups and restore procedures for critical data.

3. Freqtrade Integration

• Leverages Freqtrade for both backtesting and live trading.

• Custom adapters support:

  • Data cleaning and transformation.

  • Indicator calculations.

  • Order execution and management.