THE GROUND MOVES THIS FAST — WHICH IS EXACTLY WHY THE ROOM IS LIVE

Forget the last year. Consider the last five days.

Monday. Meta released a 30-billion-parameter multimodal model under Apache 2.0 — free, open, and small enough to run agentic work on a single GPU. Not a research toy. A working agent, on a machine you own outright.

Tuesday. xAI opened the beta of Grok Bot. These are not chat windows. Each bot gets its own computer in the cloud, signs into your tools the way an employee would, and keeps working after you close your laptop. You message one like a colleague and it comes back with the job finished. Anyone with a Cursor Ultra seat has it right now.

Wednesday. Grok 4.6 shipped — five weeks after 4.5, same price, frontier tier. 4.7 is already expected within weeks.

Midweek. Alibaba published the open weights of Qwen3.8-Max — 2.4 trillion parameters — the first time it has ever open-sourced a flagship model. Three weeks before that, a Chinese lab released the largest open model in history at 2.8 trillion parameters. Both free. Both downloadable. Both yours to run behind your own walls, permanently.

And a countdown is running right now on Alibaba's own servers. When it hits zero, the open weights of a model you can run on the laptop in your bag go public. Free. Forever. No vendor. No terms. No one watching.

Five days. And this was not an unusual week.

Any one of these would have been the story of the year a decade ago. Now they land faster than most family offices schedule a meeting.

This is the environment. Permanently. It is why a book or a course filmed six months ago is already obsolete — and why this room is not. The masterclass, like the architecture it builds, is current to the week you attend. Whatever has just happened by November 4–5 — and something will have — is priced into the room.

Whether you run a family office today or are deciding whether to build one, you will not be taught last quarter's answer. You will build against the world as it stands that morning — and you will leave with the architecture, not the notes.

---

THE TRADITIONAL FAMILY OFFICE WAS BUILT FOR A WORLD THAT NO LONGER EXISTS

Passive. Defensive. Designed to preserve capital in stable environments — committees, custodians, quarterly reviews, compliance calendars. Structures optimized over a century for a problem that has fundamentally changed.

The most sophisticated principals in the world do not need "wealth managed." They need a sovereign intelligence infrastructure. Active. Proactive. Operating 24 hours a day. Anticipating threats before they surface. Surfacing opportunities before they're broadly visible. Compounding institutional intelligence that outlasts every individual who contributes to it.

The language of family offices has been preservation. The language of sovereign intelligence infrastructure is dominance.

The technology to build that infrastructure did not exist two years ago. It exists now. What follows is how you build it. Every layer. Every tool. Every decision. In two days in New York City. And it holds whether your family runs an established office or is building its first — the architecture is the same, and the room meets you where you are.

"A high-performing family office is not measured by how busy it is. It is measured by the quality of its decisions, the speed of its execution, the strength of its people, the clarity of its purpose, the adaptability of its systems, and the degree to which the family becomes more capable because the family office exists." — Angelo Robles

---

TWO IMMERSIVE DAYS. ZERO THEORY. ALL OPERATIONAL EDGE.

This is not a conference. Not a seminar. Not a panel about what AI might do for family offices someday. It is an operational transformation compressed into 48 hours — built around the specific decisions, workflows, and opportunities of running a sophisticated family office right now, led by people who deploy this architecture inside real family offices every day, not consultants describing tools they have never touched.

AI is not improving gradually anymore. It has undergone a phase shift, and it is still accelerating. The families who aren't in the room won't fall behind gradually. They'll fall behind the way compound interest works — slowly, then all at once.

What keeps a family office out of this room is never the money. It's the quiet voice that says we'll get to it next year — and next year is exactly where the gap becomes permanent.

Seats are strictly limited to SFO Continuity members.

[JOIN SFO CONTINUITY TO ATTEND →]

---

WHAT THE TWO DAYS ACTUALLY PROVE

The two days exist to prove something in front of you: that the way a family office operates can be rebuilt — live, in the room, with your hands on it. The sovereign architecture is not described. It is built while you watch, and then you build it.

Most people arrive believing their office is ahead. That belief has usually never been tested against anything. The two days are the test. Most walk out having concluded the thing that changes everything after: we thought we were ahead. We are not.

That conclusion is worth more than the architecture — because it is what makes you finish the job when you get home. And that is what the membership is for. The seat is the membership.

---

YOU DON'T NEED TO BE TECHNICAL. NO PRIOR AI EXPERIENCE REQUIRED. NONE.

The single thing that keeps more families out of this room than anything else is the belief that you need to understand AI before you can use it.

You don't.

I have taught this to family office principals in their 60s who had never opened ChatGPT. To next-generation family members who thought AI was someone else's job. To CIOs, CFOs, trustees, and advisors at every level of technical background — from zero to engineers. The outcome is always the same.

Every concept explained from first principles. Every tool demonstrated before it is handed to you. Every framework built around the actual decisions of operating a family office at your level. You will not be lost. You will not be left behind.

What you will be — walking out — is dangerous. In the best possible way.

Your children and next-generation family members are welcome in the room. The families who treat this as a multi-generational investment leave with something no MBA program and no inherited network can replicate.

You leave knowing how, having started, and never alone. SFO Continuity Membership →

---

WHAT YOU ACTUALLY WALK OUT WITH

Not a binder. The understanding, the resources, the first real building — and the people building alongside you.

Be clear-eyed about what two days can and cannot do, because the families in this room are too sophisticated for a promise that doesn't survive contact with reality. Forty-eight hours does not hand you a finished sovereign operation — and we will not pretend to build one for you. Every family office is different. A system someone else built for you is exactly the dependency this room exists to end. What two days does is more durable: you leave understanding it, having done it with your own hands, holding the written resources to finish it — and knowing the people building alongside you.

First, the shift everything else depends on. Almost every family office using AI today is chatting — opening a window, asking a question, closing it. It feels productive and it leaves nothing behind. The progression this room walks you through is chatting → doing — AI that executes real work — → building — an operation your office runs on, that works when you're asleep and gets stronger every week. Chatting is consumption. Building is capital formation. Everything below is how you cross that line.

The understanding — from zero. You leave actually grasping how a sovereign AI operation works: the pieces, how they fit, why each choice matters. You stop being someone who buys AI and become someone who can direct it — the only position from which any of this is safe to own.

Your hands on your own machine. Not watching a demo screen — configuring, running a local model, directing agents on your own computer, in the room, with faculty beside you. The difference between having seen it and having done it is the difference between leaving impressed and leaving capable.

The written playbooks — faculty- and member-authored. You leave with a library of working markdowns written by the faculty and by members running these operations today: the actual files, configurations, and shared experience of people who built this inside real family offices — what worked, what broke, what they'd do differently. Not generic documentation. The distilled how-we-did-it of the room, yours to build from.

Your hardware decision, made and specced. The specific machine, the memory, the storage, the cost — sized to your office, settled in the room with faculty who run these machines themselves. You go home with a shopping list, not a research project.

The full model arsenal, and how to wield it. The frontier models — Claude, GPT, Gemini, Grok — through secure pipelines that never expose your data, and the open-weight models running entirely on your own hardware, with orchestration that routes the right model to each task. Fluent in all of it. Dependent on none of it.

Your first files, begun in your own hand. The real difference between an impressive demo and an operation that runs your office is a handful of plain-text files, and you leave having started yours:

• CLAUDE.md — a plain-English instruction sheet that tells your AI how your office works: priorities, structure, rules. What makes every agent reason like a member of your team instead of a generic assistant.
• Your soul file and standing mandates — a fixed identity and standing orders: who your AI is, how your family thinks, what it should always be doing. What makes the system act continuously, like a trusted chief of staff, instead of sitting idle until asked.
• Your first Skills — procedures written down in plain text: how your office handles one recurring task, saved so it runs the same way every time and is never forgotten.
• Your MCP map — the wiring that connects your systems — CRM, documents, deal flow, entity structure — so your AI sees all of it at once.
• Your second-brain seed — the first piece of the searchable, reasoning memory of your family's judgment. The asset that compounds for generations. You leave with it begun.

The faculty and member markdowns are your references — shared experience, not turnkey templates — because every family office is different, and the only version that truly runs your office is the one authored against your office. You start that authoring in the room, with the people who have done it sitting next to you.

And the building itself, begun. Files organize your AI; applications put it to work. In the room you see, and start, how a plain-English request becomes real software: a deal screener that interrogates every opportunity against your family's actual thesis, a dashboard shaped to your exact entities, a monitor that watches a jurisdiction while you sleep.

One of these already exists and members run it today — the Family Office Investment Thesis Advisor, a high-end deal scorecard built by a member, that interviews you, adapts to your criteria, and then interrogates any deal at full depth, every time. You learn to wield it, and you begin building your own — because the wall between having an idea and having working software is gone. You describe what you want, in plain English, and tools like Claude Code build it in front of you. No developer, no six-month project.

The peers traveling the same road. Notice who built the scorecard above. Not faculty — a member. That is how this community works: two days beside principals and families at your level, solving the same problems, sharing what they built, building alongside each other. Through the membership, that becomes one of the most valuable assets a forward-looking family can hold: not a mailing list, but real relationships with the few people on earth doing exactly what you are doing.

The finished operation is built in the weeks and months that follow — which is precisely what the membership is for. And the point of all of it is not AI for its own sake. It is a family office that is materially more effective: faster decisions, tighter execution, judgment that compounds instead of retiring.

"Learning to use AI for due diligence at Angelo Robles's AI Family Office Masterclass changed how I evaluate opportunities — and has already saved me hundreds of hours." — Member, Entrepreneur, Investor, and Family Office Principal

"Angelo's AI Family Office Masterclass takes what's coming and makes it tactical — not theory, but what to build now. He challenged me, and every challenge evolved the family's thinking. He is a large part of why our office is as AI-forward as it is today. In SFO Continuity, he has built something I haven't found anywhere else in the SFO space: cutting-edge perspective on the AI-native family office, at a level you can actually implement." — Brendan Boyle, CIO, Single Family Office 

---

WHAT THIS ACTUALLY DOES FOR YOUR FAMILY OFFICE

Real Operations. Real Workflows. Real Results.

Picture a Tuesday. You have a decision to pressure-test — a deal, a manager, a jurisdiction. You don't open a chat window. You speak to your office's intelligence the way you'd brief a chief of staff — by voice, from your phone, from a car — and tell it what you need. It spins up a swarm: dozens of agents, each taking one slice — one pulls the counterparty's history, one models the downside, one scans every jurisdiction the structure touches, one drafts the memo. They work in parallel, overnight if needed, and hand you finished analysis before your morning coffee. You didn't type. You didn't wait. You directed — and the office executed while you slept.

That is Tuesday, for a family that built correctly. And in this room you learn to build it — including spinning up agent swarms on demand and running the whole operation remotely, by voice, so your sovereign intelligence is in your pocket, not chained to a desk.

The ground is moving under this too. The frontier labs have gone all-in on finance — shipping ready-to-run agents for pitchbooks, valuations, earnings review, statement audits — already in production inside the largest banks on Wall Street. Each one is assembled from a few plain pieces — skills, connectors, subagents — the exact building blocks you learn to write in this room. What the labs packaged and sold to the banks, you build for your family, on your own stack, under your own control.

AI has no AUM to protect. No legacy relationships to preserve. No ego in the status quo. Function by function:

Investment research and due diligence. Agent swarms handle the research layer autonomously. Multiple models interrogate the same opportunity from different analytical frameworks, surfacing blind spots no single analyst finds, and your final thesis is run through frontier adversaries built to find the holes before you commit capital. What consumed your best people for weeks returns in days — more thorough and more adversarial than any human team produces alone.

Macro and geopolitical intelligence. Your agents monitor global feeds, central bank communications, and legislative developments 168 hours a week — synthesizing what matters for your portfolio into a morning briefing before markets open. The signal that emerges at 3am in a timezone where no one was watching is not missed.

Portfolio and operational oversight. Custom dashboards built in hours reflect your exact entity structure. Consolidated reporting on your schedule, without manual compilation. Concentration alerts, tax-loss windows, and rebalancing signals flagged in real time — not when an analyst gets to it.

Risk and reputation surveillance. Traditional monitoring is retrospective — you learn about a counterparty's deterioration when it becomes news. Your agents run continuous surveillance across media, filings, and financial databases, identifying threats at the signal stage, when options still exist. You don't react to risk. You anticipate it.

Legal, regulatory, and tax intelligence. Your agents monitor every jurisdiction you touch and analyze every material change against your specific structure. Outside counsel remains for judgment and execution. The surveillance and first-pass impact analysis — automated.

Succession, governance, and institutional memory. The knowledge in the heads of your founders and senior advisors is the most fragile asset in private wealth — it walks out the door with them. Digital Memory Vaults, built in the room, capture every decision, deliberation, and rationale across your family's history — AI-navigable, always accessible, never lost. The traditional path produces a binder no one reads. The Vault is built in hours and compounds with every decision from that day forward.

Next-generation engagement. The families who dominate the 2040s will not merely inherit capital — they will inherit the infrastructure to orchestrate it. We don't position heirs as passive stewards. We make them Sovereign Orchestrators: leaders who command a digital workforce with the confidence of someone who built it.

This is what SFO Continuity members are building — month by month, masterclass by masterclass. 

Join the membership →

---

YOUR FAMILY'S MOST VALUABLE ASSET IS STORED IN THE ONE PLACE GUARANTEED TO FAIL

Everything your family actually knows lives in the wrong place. In the founder's head. In the senior advisor's instinct for a counterparty. In the trustee who remembers exactly why you walked away from that deal in 2009 — and what it cost to learn the lesson.

That is the asset. Not the portfolio — the thinking that built it. And it sits on human memory, on a human timeline: one retirement, one departure, one generation from gone. Every family office in history has quietly reset itself this way — each generation inherits the capital and starts over on the wisdom.

There is now a different answer. It is the most important thing you build in the room.

You build your family a second brain. Not an archive that stores what you knew — a brain that thinks with it. A memory archive waits to be searched. A second brain compiles: every new decision — a deal on the table, a structure under stress, a succession question — is reasoned against the entire accumulated judgment of the family at once. The relevant memo, the three decisions like it, the reason you broke your own rule that one time and what it cost — assembled into the analysis before you finish asking. Eighty years ago Vannevar Bush imagined exactly this machine. The models finally made it real — and it now runs on a device that fits in a drawer, inside your own walls, owned outright.

And it compounds. Every decision from the day it is built makes it sharper. Ten years in, the family that built one reasons from the distilled experience of everyone who ever sat at the table. The family that didn't still starts most conversations from scratch. You cannot buy this later — the software will be for sale forever; the compounding only ever starts the day you begin. A family that starts this year and one that starts in three years are not three years apart. They are three years of compounding judgment apart, and that gap does not close by buying the same tools late.

Done right, your heirs do not inherit a portfolio and a story about who you were. They inherit your mind on the questions that matter — searchable, askable, never lost. That is not succession planning. That is a family that stops forgetting itself.

You begin yours in the room — on your hardware, against your real history, faculty alongside you the entire time.

---

CHAT IS DEAD. THE ERA OF AGENCY IS HERE.

"Chat isn't enough. AI needs to do things." — Angelo Robles

For three years, the game was chatting: you typed, AI answered, you took it from there. That era is over. First AI knew things (2023–24). Then it reasoned through them (2024–25) — which is where most of the wealth-management world still believes the story ends. Now — Era Three, right now — AI acts. It has goals and tools. It operates your computer, coordinates with other agents, executes multi-step workflows over hours without supervision, and delivers completed work. Claude Cowork controlling your Mac from your iPhone while you're in a meeting. Perplexity Computer running a fleet of models overnight and delivering a completed diligence package before you wake. A command layer monitoring 168 hours a week and flagging what matters before anyone on your team has opened email.

Era Three arrived. Most family offices missed the announcement. August 12–13 is where you get fully current.

---

SIX AI ECOSYSTEMS. EACH WITH IRREPLACEABLE SUPERPOWERS.

The question is no longer "which AI do we use?" It is "how do we orchestrate all six — routing the right model to the right task automatically?" That is the new alpha. That is Session One, Day One.

OpenAI / ChatGPT. GPT-5.6 shipped July 9 in three levels — Sol, Terra, Luna — and Sol is now the lead orchestrator of this architecture: a million-token context, native computer use, agentic workflows across your applications. Codex turns plain language into working software; ChatGPT Work is OpenAI's head-on answer to Claude Cowork, and we give you the honest comparison in the room.

Anthropic / Claude. Opus 5 is the deep-work standard and the second seat — the adversarial counterweight to the orchestrator, holding your institutional memory in a million-token context. Sonnet 5 is the fast, high-volume workhorse. Above them sits Claude Fable 5, the Mythos-class tier — the sharpest instrument any lab offers, summoned deliberately for the single hardest decision, then stepped back down from. And never trusted with the family's crown jewels, for reasons we cover in the room.

Google / Gemini. Gemini 3.1 Pro is the multimodal weapon: pitch videos, property images, financial statements, and legal filings read together, in one context, at once. For due diligence across unstructured material, nothing else takes in the whole picture in a single pass.

xAI / Grok. Grok 4.6 delivers Opus-class reasoning at an affordable price — and it is the only frontier model natively wired into live market data and the social layer, where a counterparty's trouble surfaces days before it reaches a filing. A different information clock than everything else in the stack.

Meta / Muse Spark. The return nobody priced in: Muse Spark 1.1 is the first model Meta has ever charged for — built for agentic tool use at roughly a quarter of top-tier pricing, speaking both the OpenAI and Anthropic API formats. The cheapest uncorrelated fourth opinion in the stack.

Perplexity. No longer a search engine. Perplexity Computer orchestrates a fleet of frontier models on autonomous workflows, and its Model Council puts several in parallel on your most consequential decisions — forcing independent conclusions and showing exactly where they agree and where they break apart. Two hundred dollars a month.

---

AND THE FRONTIER JUST PROVED IT CAN BE SWITCHED OFF — FOR EVERYONE, OVERNIGHT

In early June, Anthropic released Claude Fable 5 — the most powerful model the lab makes available to anyone. For a few days, you could reach it. Then a U.S. export-control directive ordered it restricted, and the lab did the only thing it could: it switched Fable 5 off — for every customer on earth, overnight, without recourse. No amount of money reopened it.

It stayed dark for more than two weeks. On July 1 it came back — worldwide, in a day — but capped, carrying a mandatory data-retention window that no customer, on any tier, at any price, can escape. The frontier vanished by someone else's decision, returned by someone else's decision, and now operates on someone else's terms. In no direction did the customer get a vote. Not the largest single family office on earth. Not you.

And it was not only Anthropic. OpenAI's GPT-5.6 Sol spent that same window behind a government safety review of its own, shipping only when the review cleared, on a calendar the lab did not set. Two of the most capable models on earth, gated the same month. That is the shape of the frontier now.

Watch what the families running sovereign architecture did with that month. When Fable went dark, they lost nothing — their floor, the open weights inside their own walls, never blinked. When Sol shipped at half of Fable's price, they moved their lead model to it in an afternoon. No migration project. No committee. No loyalty. In this architecture, no model is an allegiance — every one is an employee in a chair, and the chair belongs to the family.

The ceiling — whatever ships next — is access someone else grants, revokes, and re-prices on a calendar you do not set. The floor — the open weights inside your walls — answers to no release schedule and no safety review but yours.

Stay current as AI evolves. That's what the membership is for. SFO Continuity Membership →

---

THE SOVEREIGN INTELLIGENCE MANDATE

Stop Bleeding IP. Start Building Your Fortress.

Every prompt your team enters into a commercial AI without a sovereign architecture is an intelligence transfer. Your investment theses. Your tax strategies. Your succession architecture. Your deal terms. Processed, stored, and potentially training the models your competitors will use against you tomorrow.

Don't take it only from me. Alex Karp built Palantir on the thesis that serious institutions must own their compute, their models, and their data — that real safety means knowing you "own the means of production, and it's not being transferred to someone else." Not a vendor's promise. Ownership.

A frontier lab is not your advisor. It is a counterparty whose next product is built from what its customers teach it — you saw the proof above: the finance agents the largest banks now run were assembled from what those banks taught the frontier. So turn the questions on your own office. Who owns your data? Where is it cached? Are your prompts secure — or training material? If you can't answer cleanly, you don't have an AI strategy. You have an intellectual-property leak with a monthly invoice.

And the leak is only half the exposure. You cannot rent intelligence from the same place that rents it to your competitor. Even with perfect security, every office in your peer set feeding the same frontier models converges on the same analyses, the same conclusions, the same quarter. The rented model is the great equalizer — everyone paying a premium for the privilege of becoming more alike. The families in this room build the opposite: a sovereign corpus no other family holds, judgment no other family encoded. Everyone has the frontier. Only your family has your corpus. That is the edge.

The question is binary: Sovereign Intelligence — or Leaked IP. 

"True Sovereignty means owning your AI infrastructure like you own your family's gold." — Angelo Robles

---

SOVEREIGN IS NOT ONE THING. IT'S SIX.

Most family offices thinking about AI sovereignty fixate on data privacy — rightly. But data leakage is only the first layer. Miss any of the six and you are still exposed. Command all six and you have built something that cannot be taken from you.

01 — Data Sovereignty. Every prompt entered into a commercial AI is transmitted to servers you don't own, under retention policies written by lawyers whose clients are those companies, not you. Sovereign architecture routes that intelligence through locally hosted models on hardware you control. The data never leaves. The models never phone home.

02 — Terms Sovereignty. Commercial vendors update their Terms of Service unilaterally. What is permissible today can shift overnight, and you have no recourse — you clicked accept. With sovereign infrastructure, you set the terms. Permanently. There is no counterparty with the power to rewrite your operating environment.

03 — Operational Sovereignty. Vendors go down, throttle access, pivot their offerings — documented events, not hypotheticals. Your Vault runs whether or not a San Francisco server farm is having a bad day. No usage limits, no rate caps, no silent model swaps changing your most sensitive workflows without your knowledge.

04 — Audit Sovereignty. Regulators, trustees, and future generations will one day ask: what AI touched this decision? With sovereign architecture, every query and output is logged in infrastructure you control — full provenance, auditable by whoever you authorize. With commercial AI, you have a vendor's approximation of an audit trail, subject to their interests.

05 — Model Sovereignty. Vendors deprecate models; if your workflows are built on one that disappears, you rebuild on their timeline. Sovereign infrastructure gives you model continuity: you choose what runs, when it updates, and whether it updates at all. The decision is yours.

06 — Cost Sovereignty. You are renting intelligence at prices set by others, designed to capture value as your dependency deepens. The sovereign family office pays once for hardware and runs models at effectively zero marginal cost per query. No per-token rent. No subscription that can be repriced.

The sovereign family office does not ask permission to protect its intelligence. It builds infrastructure that makes the question irrelevant.

---

THE SOVEREIGN STACK: BUILT FOR YOUR FAMILY OFFICE. IN THE ROOM. IN TWO DAYS.

Two years ago, running frontier-capable models on local hardware — privately, at family-office cost — was not possible. That changed. The window to build sovereign infrastructure before it becomes table stakes is open now. Not indefinitely.

Layer One: The Cognitive Vault — a device you own, not a brand you rent. The Vault is a category, not a logo: a small machine in a cabinet you control, running frontier-capable open models with zero cloud exposure. There are three credible ways to build one, and we configure the right one for your family in the room — by ecosystem, by budget, by who maintains it.

The Apple path — the default for most families. The Mac Mini remains the simplest Vault: silent, briefcase-sized, built for continuous operation, at home in any office already living in Apple — with the M5 generation arriving across the back half of 2026. One honest note: the global memory shortage has reshaped this lineup in real time — Apple raised prices in June and capped high-memory configurations — which makes the timing argument concrete. The window to build affordably is open now, not indefinitely.

NVIDIA DGX Spark — the Vault for the NVIDIA world. A Grace Blackwell supercomputer the size of a Mac Mini, 128 gigabytes of unified memory, the full NVIDIA stack preinstalled, running open models up to 200 billion parameters on your desk — and two of them link with a single cable into one 256-gigabyte pool.

AMD Ryzen AI Max+ "Strix Halo" — the most capacity for the least money. The first x86 chip to put 128 gigabytes of unified memory on a single package, in a mini PC running Linux or Windows, at the lowest entry price of the three. For the family that wants the largest sovereign models at the value end.

Three doors. One principle. The intelligence runs on hardware you own, on open weights answering to no vendor's release schedule, and nothing your family thinks ever leaves the building. Which door is right for you is settled in the room — not from a sales page, including this one.

On your Vault, your sovereign arsenal. The everyday sovereigns — Gemma 4 (Google's open weights, frontier-level reasoning small enough to run fast and local), DeepSeek V4 (open frontier reasoning under MIT license), Qwen 3.6 (native multimodal, a hundred-plus languages, built for multi-jurisdictional families), gpt-oss (OpenAI's own weights in your hands, Apache 2.0), and Nemotron 3 Nano (NVIDIA's compact American reasoner, tuned to fly on its own hardware) — carry the bulk of real family-office work on a Mac Mini-class machine.

Above them, the giants: GLM-5.2 (among the highest-ranked open models in the world, a genuine peer to the closed frontier), Nemotron 3 Ultra (the 550-billion-parameter American heavyweight), and Kimi K2.6 (a trillion-parameter agentic specialist). Be clear what giants mean in metal: these are server-class systems — not the Mini, not a maxed Studio. Running one clean means real GPU iron, or the honest alternative most families choose: the overnight batch run, for the heaviest questions only. We size that honestly in the room — including whether you need the giant at all.

The American lane now has a flagship: Inkling, from Mira Murati's Thinking Machines Lab — at 975 billion parameters, the largest American open weights ever released, Apache 2.0, multimodal, downloadable from day one. When provenance matters and you have the hardware, this is the U.S. flagship.

And Kimi K3 — the largest open-weight model ever released: 2.8 trillion parameters, shipped as public weights July 27. Genuinely competitive with the closed frontier flagships, as a download you own — and far beyond any hardware a family office runs today. Which makes this architecture's standing doctrine a live instruction: download and archive the weights on storage you own, because owning the artifact is sovereignty, even years before you own the machine to run it. Own it this quarter. Run it when the hardware curve catches up.

Why a Chinese model on your Vault never touches China. Every model above runs as open weights: you download the file once from a public registry and it executes entirely on your hardware. You never call the lab's API. A downloaded weight is inert with respect to its maker — no telemetry, no visibility, no way in. The exposure was never the model. It was the hosted API, and the sovereign architecture does not call one. The standing rule is absolute: a Chinese-hosted endpoint is never touched — for any model, at any ranking. What remains is provenance — whose research produced the weights — which is why the arsenal is deliberately mixed: the Chinese frontier for raw capability, the American options for the mandates where provenance is part of the decision. Owning the stack means you make that call yourself, model by model. We build that map with you in the room.

Layer Two: Frontier reasoning — secure pipelines, zero DNA exposure. Your Vault connects through encrypted pipelines to the frontier models — but your Institutional DNA never leaves. Before any query goes out, your sovereign layer strips the identifiers: the family name, the entity structure, the holding, the counterparty. What leaves is the question without the fingerprint. The frontier applies its full reasoning and returns the intelligence; your Vault reassembles the answer inside your walls, against your full proprietary data. Sovereign in. Intelligence back. Your DNA never exposed.

Layer Three: MCP — the architecture that makes everything one. Before MCP, every AI tool was an island. MCP is the universal standard that connects all of it into a single sovereign system: deal flow feeding due diligence agents, CRM feeding relationship context into every analysis, regulatory monitoring wired to your entity structure, your document vault searchable by every agent at once. We architect your MCP integration in the room, built to your specific infrastructure.

Layer Four: The autonomous workforce — 168 hours a week. A 40-hour work week is a sprint. Continuous operation is the new standard: risk and reputation surveillance catching a counterparty's deterioration at the signal stage, weeks before it is news; research swarms connecting macro signals to portfolio implications while the world sleeps; legal and regulatory agents tracking every jurisdiction in real time; perimeter defense against the intelligence threats that come with being a family worth targeting; capital optimization flagging the window that opens at 2am and closes before morning. Your people's time freed entirely for what only human judgment can do.

The phase transition your family office was not built for. Agrarian era: control of land. Industrial era: control of capital. Financial era: control of information asymmetry. AI era: speed and quality of decision-making under uncertainty. Each transition, the families who moved early dominated. Information is now a commodity — your competitors have the same data and the same feeds. The only remaining alpha is Decision Velocity: processing global complexity at machine speed, validating through adversarial multi-model analysis, executing before the window closes. August 12–13 is where you rebuild for it.

"The SFO of the future isn't just a smaller version of today's model. It is a sovereign fortress of intelligence that produces superior results through elegant efficiency." — Angelo Robles

---

THE TOOLS THAT DIDN'T EXIST LAST YEAR

OpenClaw. Hermes. Perplexity Computer. Claude Cowork. Claude Code. Live. In the Room.

OpenClaw — your sovereign command layer. Soul file. Heartbeat. Standing mandates. Your family's AI knows how your family thinks, runs continuously without being prompted, and delivers completed work while you sleep — one agent when the task is simple, five coordinated agents when the problem demands it. Every morning: a full intelligence briefing assembled overnight — positions, macro developments, risk alerts, opportunities — delivered to your phone through WhatsApp or Signal before your first coffee. Every task: send a message from anywhere; your digital workforce executes, context-aware, no briefing required. OpenClaw routes each task to the right model automatically — research to Perplexity, financial analysis to Opus, real-time intelligence to Grok, coding to Claude Code. We don't demo it. You see it running, and you leave knowing exactly what your office has been missing.

Hermes — the open agent harness you own outright. OpenClaw gives you the command layer; Hermes is what you run when you want that layer answering to no one but you. The open-weight agent framework — tools, memory, multi-step execution — running on your hardware, none of it phoning home. For the family's most sensitive work, the deliberations that can never touch a cloud, this is where it runs. And as of this month you no longer sit at the Vault to drive it: Hermes ships native apps for Mac, Windows, and phone — but the brain does not move. The agent stays resident on hardware you own, behind your private network, while you reach it from a laptop or a message from a car between meetings. Start a thread at your desk, finish it from a plane, never re-explain a thing. The intelligence stays home. Only the control travels. It is the cloud convenience you were told required the cloud — running entirely on what you own.

Perplexity Computer — 19 models, one autonomous operator. You define a goal; it breaks the work into tasks, routes each to the optimal model among 19 frontier systems, and executes until finished work is delivered. Model Council runs frontier models in parallel on your most consequential decisions and shows you exactly where they agree and where they diverge. Two hundred dollars a month.

Claude Cowork — your office, automated. Agentic AI for knowledge work: you define the goal; Cowork connects to your files, calendar, communications, and financial systems and delivers finished, formatted, ready-to-review output. Recurring tasks automated on your schedule. Dispatch assigns work from your iPhone; you return to it finished on your desktop. And when Cowork lacks a connector, Claude now uses your Mac directly — opening applications, navigating browsers, filling spreadsheets — anything you can do at your desk, done while you're away from it.

Claude Code — your own family office software, in hours, without a developer. Describe what you want; Claude Code builds it. Custom portfolio dashboards for your exact entities. Deal flow intake and scoring built to how you actually work. Governance tools, risk monitors, Digital Memory Vaults. No contract, no six-month timeline. We build live in the room, on real family office use cases — and you begin building your own, with the faculty and member playbooks in hand.

Bolt — the on-ramp. You describe an app in plain English and watch it become a live application in your browser, nothing installed, no code written — the fastest way to feel what building software now costs, and to find out whether an idea deserves real engineering before a dollar is spent. It also teaches the sovereignty rule by contrast: Bolt runs in someone else's cloud — perfect for the prototype, wrong for the family's real corpus. Prototype in Bolt to find the shape; rebuild on the Vault with Claude Code the moment it touches anything sensitive. Rent the rails to discover the idea. Own the reasoning that runs on real data.

Dynamic Workflows — orchestration, built into the frontier. Anthropic moved orchestration into the core: hand Claude a complex job and it plans the work, fans it across hundreds of subagents running in parallel, has them challenge each other until they converge, and returns one verified answer that has already survived its own adversaries. The Gladiator Method as native infrastructure. The sovereignty discipline still governs it — it runs in Anthropic's cloud, so it earns its place on bounded, sanitized work, never the family's most sensitive deliberations. The capability is extraordinary. The deployment surface is the decision. We draw that line in the room.

Granola — the meeting that never gets lost. Every meeting — in person or on any platform — captured, transcribed, and converted into structured intelligence: decisions reached, commitments made, next steps by owner. The transcript routes into the sovereign pipeline via MCP, filing into the right deal record within minutes. For a family building a Digital Memory Vault, Granola is the layer that captures how decisions get made — not just what was decided.

Paperclip — the governance layer that prevents AI chaos. Everyone who deploys agents at scale discovers the same problem within weeks: duplicated work, spiraling API costs, vanishing accountability. Paperclip sits above the agent layer — the org chart and budget authority for your digital workforce: mission-level objectives, defined roles, enforced compute budgets, and a complete audit trail of every agent decision. Install it from the beginning, or learn the lesson at cost.

---

ABOUT ANGELO ROBLES

Angelo Robles is the Founder & CEO of SFO Continuity and Billionaire Mastermind, family office strategist to the world's wealthiest families, and a global authority on the super-rich. Over two decades, he has worked within the ecosystem of more than 100 billionaire families — observing up close how they think, decide, and build enduring wealth systems. He previously founded and exited Family Office Association, one of the world's most recognized family office networks.

When Angelo demonstrates something in the room, it is because he runs it himself. Every tool. Every workflow. He is not teaching theory — he is transferring an operational edge built over years, including inside a 700-year-old European dynasty he advises on sovereign AI architecture today.

"The majority of family offices today are built for a pre-AI world. The future demands new intelligence, new structures — and new leaders. The families who transform now will rule the next century. Those who wait will spend it begging for access to what others built." — Angelo Robles

---

THE URGENCY IS REAL. THE GAP IS WIDENING. EVERY. SINGLE. DAY.

The families building sovereign AI systems today are developing operational fluency, proprietary workflows, and institutional memory that cannot be replicated by purchasing better software later. The intelligence compounds. The workflows compound. Late movers never catch up simply by adopting the same tools at a later date.

A hard question for the operators reading this: if you've built your career on knowledge and experience — how do you win in twenty-four months if you don't adapt?

Here is the most practical sentence on this entire page: nothing available to you in your career right now will have a more immediate impact than becoming genuinely good with AI in the next thirty days. Nothing is close. August 12–13 is the thirty days, compressed into two.

Experience is still your edge. But experience that refuses AI isn't an edge anymore. It's a liability. Don't be arrogant. Adapt. The operators who win the next decade are the ones senior enough to know what mattered — and humble enough to learn the new tools. Both.

DON'T FIGHT TECH — IT ALWAYS WINS.

The question is not whether AI will reshape family offices. That question has been answered. The question is whether your family will be among the architects — or among those who had the chance to lead and chose to follow.

---

JOIN THE SFO CONTINUITY COMMUNITY

The Membership That Makes This Possible — And Keeps You Current

The masterclass is two days. The SFO Continuity membership is 365 days a year — and the masterclass is one of many flagship benefits included. AI is not a project with an end date. It is a permanent competitive commitment, and the families who dominate the next decade will be the ones who stayed current — month by month, tool by tool, breakthrough by breakthrough.

What membership includes:

All Masterclasses — November 4-5 in New York City, plus every masterclass Angelo hosts over the following twelve months. When the landscape shifts again — and it will, within months — you are in the room.

Ongoing Access to Angelo and the Faculty Network — genuine ongoing access to the practitioners furthest ahead, as new tools emerge.

The SFO Continuity Intelligence Network — the community of AI-forward family offices moving together at the frontier: sharing operational experience, comparing what works, building alongside each other.

Member-Exclusive Resources — both of Angelo's living books — The AI-Native Single Family Office and 10 Things Every Family Office Should Be Doing Now in AI — updated every few weeks as the landscape moves, plus the full library of recordings, frameworks, and deployment guides.

Priority Access to Emerging Tools — private briefings and early deployment guidance on new AI systems as they reach family office readiness.

The Network — real relationships with the AI-forward principals, advisors, and operators building what the future of private wealth looks like. The peer asset that compounds over time.

$10,000. Twelve months. Everything. The first 100 members join at that rate and lock it for life; after 100, membership is $15,000. That is the entire structure — no tiers, no upsells.

This is not a conference membership or a content subscription. It is the operating system for the families who intend to lead.

[Join Now — Pay Securely]

[Schedule a Private Briefing]

---

NOVEMBER 4-5. NEW YORK CITY. EVERYTHING CHANGES.

Two days will not turn you into an AI engineer. That is not what this is. What two days will do — with Angelo and this faculty, in this room — is give you the complete operational blueprint, hands-on experience with every tool, and the practical knowledge to go home and execute with confidence.

You leave understanding the Sovereign Stack at a level you can act on — the Vault, the local models, the frontier pipelines, the MCP architecture — as something you have personally touched and run. You leave having operated OpenClaw, Perplexity Computer, Claude Cowork, and Claude Code yourself, on real family office scenarios, faculty alongside you the entire time. You leave with the faculty and member playbooks, your first files begun, and a deployment roadmap built for your office — so the week after you return home is execution, not evaluation. And you leave as an SFO Continuity member, with Angelo, the faculty, and the peer network behind you for the twelve months that follow. The masterclass is the beginning of the transformation, not the entirety of it.

This is not an event you attend. It is an operational turning point you cross.

The families who control the 2040s are being decided right now. Your great-grandchildren will either inherit a dynasty that shapes the world — or they will serve families who had the courage to act when it mattered.

The choice is yours. The time is now.

Email Angelo directly: [email protected] 

Call Angelo directly: 203-570-2898

SECURE YOUR SEAT & JOIN SFO CONTINUITY

November 4-5 | New York City | 12-Month Membership angelorobles.com/membership

P.S. You have multiple Einsteins on call, twenty-four hours a day, and the sovereign architecture to keep everything they touch inside your walls. The only question left is whether you build it — or watch someone else build it first. Your children will inherit the consequences of what you decide today. The seat is available. The decision is yours.

---

Sovereign or Obsolete. The AI Family Office Masterclass | November 4-5, 2026 | New York City The seat is available. The membership is open. The only question is whether your family is in the room.