Why InvestOS

InvestOS, our open-source portfolio engineering and optimization engine, is built to help institutional investors turn research signals into real, investable portfolios with clarity, control, and discipline.

Most portfolio tools break down when real-world constraints, transaction costs, and evolving signals are introduced. InvestOS is designed from the ground up to handle those realities while remaining transparent, flexible, and production-ready.

The Core Idea

  • Portfolio engineering, not black boxes: InvestOS provides explicit, controllable portfolio construction logic rather than opaque optimizers.
  • Signals in, portfolios out: combine alpha signals, risk models, constraints, and costs into a single, coherent optimization framework.
  • Built for real constraints: handle turnover limits, position bounds, sector and factor exposures, transaction costs, and liquidity assumptions directly.
  • Research and production aligned: the same code paths can be used in research, backtesting, and live portfolio construction.
  • Solver-flexible by design: supports both open and closed-source optimizers via cvxpy, letting teams choose the solver that best fits their performance, licensing, and infrastructure needs.

What You Can Do With It

  • Build investable portfolios: translate factor and model outputs into portfolios that respect real-world constraints and trading realities.
  • Test portfolio decisions explicitly: understand how constraints, costs, and risk targets affect outcomes rather than relying on heuristics.
  • Run systematic strategies at scale: construct and rebalance portfolios consistently across universes, mandates, and frequencies.
  • Integrate seamlessly: use InvestOS alongside ForecastOS FeatureHub and Hivemind, or plug in your own signals and risk models.

Why It Matters Today

  • Alpha lives or dies at the portfolio layer. Even strong signals can fail if portfolio construction is brittle or misaligned with reality.
  • Markets are more constrained. Liquidity, crowding, and cost awareness are essential for sustainable performance.
  • Institutions need transparency. InvestOS makes portfolio decisions auditable, explainable, and defensible to stakeholders.

Open Source by Design

  • MIT licensed: do whatever you want with it. Use it, modify it, extend it, or embed it in proprietary systems without restriction.
  • Fully open-source: InvestOS is available on GitHub, allowing teams to inspect, extend, and trust the portfolio construction logic.
  • Model-agnostic: bring your own signals, features, and risk models without being locked into proprietary assumptions.
  • Production-ready foundations: designed for correctness, reproducibility, and long-term maintainability.

Next: InvestOS External Guides

Let's explore the InvestOS External Guides next.