What is PWOS
What is PWOS?
PWOS is the Protocol Wealth Operating System — the in-house system our advisory practice runs on. It governs how AI is used so client data stays out of the models, a human fiduciary approves every client-facing output, and every decision is recorded for reconstruction.
Most firms add AI and compliance on top of off-the-shelf tools. Protocol Wealth built its own operating system instead, so that the rules a fiduciary has to follow are enforced by the system itself rather than by a policy memo. Our advisor and client surfaces — the advisor workspace and the client portal — both run on PWOS.
The organizing idea behind PWOS is one sentence: each layer exists to make the layer above it trustworthy.
Why it exists
Why we built it
AI made analysis cheap. A client with an AI subscription can now reach most of what an advisor knows. What they still cannot get is an advisor's judgment under ambiguity, accountability as a fiduciary, and the work of connecting a recommendation to the reality of their own life. We wanted AI as an instrument the advisor controls — governed and recorded — not an autonomous black box.
Doing that inside a regulated investment-advisory practice means three things have to be true by construction, not by good intentions: sensitive client information should never reach the models, a human should approve anything a client sees, and it should always be possible to reconstruct what happened. Off-the-shelf tools do not give you those guarantees. So we built the substrate that does.
How it works
The layered model
PWOS is a stack. Reading it from the bottom up is the fastest way to understand why it is not just another AI application: each layer exists to make the layer above it trustworthy.
Decides what an AI or a person may do, and who signed off — a human-in-the-loop review on every client-facing output, a two-step confirmation on actions that change client-visible state, a partner sign-off process, and a compliance-officer- governed exception system. Makes the practice trustworthy.
How a recommendation is produced — our systematic investment framework (PWAF, built on the EMF methodology): classify the market environment, score each asset, size the position, review continuously. Makes advice trustworthy.
The evidence a decision stands on — the regime and scoring engine (open-sourced as nexus-core), market, macro, and on-chain data, and planning tools. Grounds the decision layer in evidence.
What makes all of the above examiner-defensible — keeping client information out of the models, non-custodial custody (we never hold client private keys), an immutable audit trail, and an AI provider (Anthropic) operated under a Zero Data Retention agreement. Makes governance, decisions, and research defensible.
AI governance
How AI is governed
Our posture is human-in-the-loop: AI assists advisor work product; no client-facing AI is autonomous. In practice:
- AI drafts; a human emits. The AI produces a draft; an advisor who knows the client reviews, edits, and sends it. There is no automatic send path.
- Client data is kept out of the models. Sensitive client information is excluded from what the AI sees — by construction, rather than scrubbed after the fact.
- Risky actions take two steps. Anything that would change client-visible state requires an explicit confirmation step the person has to approve.
- Everything is recorded. Each decision and AI action is written to an immutable audit trail that captures who and what produced a result, so it can be reconstructed end to end.
- Exceptions are governed. The narrow cases where the default controls would otherwise block a legitimate task are handled through a compliance-officer-approved, fully recorded exception process.
More on the principles behind this is at our AI Principles; the security posture is at /security.
Honest status
What is live today — and what is not yet a product
We try to be precise about what is shipping versus what is still being built.
Live today
- Protocol Wealth's advisory practice runs on PWOS.
- The governance, decision, research, and trust layers above.
- The investment framework engine, open-sourced as nexus-core.
- The reusable compliance primitives, published as open source.
In development — not yet a product
- Licensing PWOS to other registered investment advisers.
- AI deliverable assistants that draft long-form analyses for advisor review.
- A standalone governance/credential-layer offering for institutions.
These are described as direction, not as offerings you can buy today.
The category
A credential layer for AI
The category PWOS belongs to is sometimes called the "credential layer for AI," or governed decision infrastructure for AI-enabled financial advice. Protocol Wealth is the registered-investment-adviser-shaped instance of it: because we are ourselves an SEC-registered adviser, the rules that govern books-and-records, client privacy, custody, and advertising are first-class concerns in the architecture rather than a governance layer left for someone else to add.
That is a description of how we operate. It is not a performance claim, and it makes no comparison to any other firm.
Because the research and scoring tools in the layers above are published as open source, they are also reachable over the open Model Context Protocol (MCP) — the same open standard a number of AI clients and agent platforms speak, among them Claude, GPT, Gemini, and SmythOS. That means an MCP-compatible client can call our public, read-only research tools directly, and it is the direction in which a compliance and credential layer like PWOS is designed to be called or wrapped by an agent platform.
This describes compatibility with an open standard, not an integration, a partnership, or an endorsement of any third party. Naming a platform is illustrative of what MCP compatibility means; it does not imply that platform uses PWOS or that PWOS uses it.
See for yourself
Where to look
Open source →
The reusable primitives, published so you can inspect the approach.
What is EMF? →
The investment framework that lives in the decision layer.
Agents →
How we think about AI as a credential layer.
How we work →
Engineering-process transparency for the substrate.
Security →
The posture behind the trust layer.
AI Principles →
The principles that govern how we use AI.
Last updated: June 5, 2026. Protocol Wealth, LLC is an SEC-registered investment adviser (CRD #335298). See our Form ADV for authoritative regulatory disclosures. Registration does not imply a particular level of skill or training; advisory services are provided only under a signed advisory agreement.
This page describes how Protocol Wealth operates. It is general information, not a performance claim, a guarantee of any outcome, or a comparison to any other firm, and it is not investment, tax, or legal advice. References to capabilities still in development describe direction, not products available for purchase.
All investments involve risk, including the potential loss of principal. Past performance does not guarantee future results.