Warren

Executive Summary

Warren is an automated backtesting and algorithmic trading platform designed to enable non-technical users to develop, test, deploy, and monetize trading strategies without programming expertise.

Problem Statement

Manual backtesting is time-consuming and error-prone, requiring hours of meticulous work in spreadsheets. Existing automated backtesting platforms require significant programming expertise and development time, creating a barrier to entry for traders without software engineering skills.

Solution Overview

Warren provides a no-code platform for designing trading strategies, with comprehensive backtesting against historical market data. Once validated, users can deploy an integrated trading bot for live market execution. As the platform matures, a marketplace will enable users to monetize their profitable algorithms.

Architecture

Warren uses a model-driven microservice architecture deployed on Railway. Services communicate via FastAPI, and each service contains a Pydantic model of the data it processes that is accessible by other services via the Backtest Orchestrator.

The Backtesting Dashboard initializes its input fields by requesting the Core Backtesting Settings model, a list of available signal, stop loss, and take profit algorithms, and the risk management data model.

The Backtest Orchestrator retrieves data models from the respective services to understand the expected input formats for future processing. Through the Market Data Manager, the Backtest Orchestrator validates the user-specified date range for the backtest. Once the user selects algorithms and configures core and risk management settings, the Backtest Orchestrator orchestrates the backtest execution and returns the results to the Backtest Dashboard for visualization.

Monetization Strategy

Platform access will be free, but users will pay for value-added services. Revenue streams include algorithm and indicator marketplace commissions from sales and leases in the user-driven marketplace, tiered subscriptions for historical market data access based on the amount of data available for backtesting, and flexible backtesting pricing with monthly subscription tiers or pay-per-use credits for compute-intensive runs.

Additional revenue opportunities include tiered API access with higher rate limits and programmatic trading capabilities for professional traders, white-label licensing for brokerages and institutions with custom branding, and professional services such as custom algorithm development, onboarding, training, and strategy consulting.

Technical Challenges

As a solo project, Warren requires expertise across multiple complex systems including frontend development, backend architecture, database design, and financial market integration.

Integrating AI into algorithmic development presents a significant challenge. The system must accurately translate a user's vision, intention, and nuanced trading concepts into precise programmatic definitions. This translation requires the LLM to understand both the strategic intent behind trading decisions and the technical requirements for implementing them correctly.

Establishing a market ecosystem for leasing algorithms and indicators requires building the infrastructure for algorithm discovery, validation, pricing, licensing agreements, and transaction processing, while ensuring security and trust between algorithm creators and users.

Current Status

Warren has been deployed to production on Railway with a test database containing limited historical data. The microservices are communicating successfully, and the backtesting engine is nearing completion.