Built with Mimir. Runs on Mimir.

Two products from Feature Factory — one you'll install, one you'll learn from.

Mimir (Playbooks for AI-assisted engineering)

Feature Factory's flagship. A local MCP server that hands your team's playbooks, rules and skills to any AI IDE that speaks MCP — so every assistant in your team reads from the same book.

Mimir activity view with Markdown guidance, linked skill and artifacts

Activity view

Rich-Markdown guidance, linked Skill on the right, linked Artifacts below.

Mimir MCP tools listed in an AI IDE

MCP tools, wherever you code

Mimir's MCP tools loaded in Cursor: create_playbook, export_workflow_to_local, get_activity…

IDE rule file exported from Mimir playbook

Rules that reach into the editor

.windsurf/rules/pytest.mdc with YAML front-matter and alwaysApply: true.

Phainestai (ML/AI-driven investment platform)

A quantitative portfolio engine for running ML strategies across global markets. Data pipelines, strategy research, backtests, live execution and track-record — one coherent system, built the Feature Factory way.

Phainestai portfolio and equity curve

Portfolio & equity curve

Live PnL, exposure and drawdown across the active book.

Phainestai strategy research and backtest

Strategy research & backtest

Walk-forward backtest with Sharpe, hit rate and turnover panel-by-panel.

Phainestai live execution across venues

Live execution

Mercury executor routing orders across venues with latency and fill-rate telemetry.