Skip to main content

Extend FundPro with AI

Use AI Researcher to run backtests and work with Merlin and Q-API. This page covers custom development when you want to extend those workflows or build an integration.

Develop with your research context​

Describe the intended behavior, identify the relevant strategy and saved results, and ask your coding agent to inspect the existing implementation before changing it. Keep reusable custom code in ~/fundpro/user_code and campaign-specific work with its research artifacts.

For example:

Use AI Researcher to inspect the saved results from [campaign]. Help me
create a reusable analysis in my user_code directory that compares
[metric] across those runs. Use the existing FundPro interfaces and
explain how to run the analysis again.

The source, type information, and examples in the following resources can help your agent understand the interfaces:

Supplementary AI assistants​

The MesoSim AI Assistant and FundPro AI Assistant remain available for strategy and product conversations. Use AI Researcher when you want to act on your connected FundPro instance and retain research artifacts in your workspace.

Codex​

For FundPro research, install the MesoSim and AI Researcher plugins using the setup guide. For custom code, provide the relevant files and ask for the implementation and validation needed for your change. See the official plugin documentation for current application guidance.

Claude​

Follow the AI Researcher setup guide for the supported desktop installation. When extending Merlin, give your coding agent the source and a concrete development objective; consult the Merlin developer documentation for the environment and validation steps.

Gemini in notebooks​

For custom notebook work, Colab's Gemini integration can help explain or modify Python cells. See Google Colab for notebook-specific instructions.