Discover tools for AI investment research

Payment-method-agnostic protocol for agent payment authorization

Its additive model fits nonlinear trends with yearly, weekly, and daily seasonality plus holiday effects.

Coinbase AgentKit provides wallet and on-chain action tools rather than a trading strategy

Supports live execution

Available in several programming languages, it can retrieve market data and serve as a base for trading algorithms, backtests, and bots.

Supports live execution

Built on CVXPY and integrated with Pandas, it supports multiple risk objectives and portfolio-construction methods.

It calculates weighted average entry, break-even price, margin, and liquidation level per order, including liquidation timing relative to the next order.

The official repository presents research workflows and modules for market data, portfolios, backtesting, and quantitative research.

It groups observations into chosen time intervals to aggregate market or other time-series data consistently.

It supports strategy development, derivatives structuring and analysis, risk management, and statistical-data applications.

Develop crypto strategies with tools for historical backtesting, result plots, money management, and optimization, with operation through a web interface or Telegram. You can first assess an idea against historical data, then review the documented options for simulation or live setup.

Who it suitsPython users who want to write and evaluate crypto strategies.

Coding required: Coding required

Supports live execution

Eliza framework for hosting agent workflows; trading depends on installed integrations

The repository collects experimental answers to chapter exercises for readers studying and reproducing quantitative methods.

It provides return and moving-average analysis alongside portfolio-optimization functions, including efficient-frontier exploration.

It creates cryptocurrency datasets, trains an up-or-down prediction model, and runs AutoTrader simulations using data collected through Coinbase’s API.

The repository is documentation-first, collecting engineering, strategy, risk, compliance, and infrastructure playbooks rather than runnable trading code.

The project describes LSTM-based stock-behavior prediction and provides a graphical interface.

It generates features using statistical, signal-processing, and time-series methods, then selects relevant features with statistical tests.

It covers bonds, equities, FX, interest-rate, and credit products with inspectable and extensible pricing models.

It provides a visual interface and supports manual instances plus workflows such as backtesting, Telegram, and TradingView.

Supports live execution

The repository describes itself as an implementation of Freqtrade for experimenting with different machine-learning methods.

Supports live execution