For family offices
A dedicated AI environment that helps your office search estate documents, summarize investments, answer tax questions, prepare for meetings, and run household operations — without exposing private family information to outside AI providers.
For single- and multi-family offices managing estate plans, trusts, private investments, tax records, philanthropy, and household staff under strict confidentiality.
The problem
Estate plans, trust agreements, LPAs, side letters, K-1s, tax records, advisor correspondence, household budgets. The information is exactly the kind a family office cannot place into a chat window — and exactly the kind a small staff cannot keep in their heads.
Trapped in
Eaten by manual work
A private AI computer gives a small office the leverage of a much larger one — without exposing the family to a vendor that could change its terms next quarter.
What you build
Indexed across estate, investment, tax, philanthropic, and household documents. Reads what you allow. Cites every answer back to a source you can verify. Writes nothing without principal or chief-of-staff sign-off.
Search across the full document archive
Plain-English questions over trusts, LPAs, tax, and operations.
Trust and estate Q&A
Distribution rules, trustee powers, definitions — answered from the actual document.
Investment document summaries
LPAs, side letters, K-1s, and direct-deal memos in structured form.
Tax document review
Year-over-year comparisons across entities and partnerships.
Meeting and briefing preparation
Family, trustee, advisor, and philanthropy committee meetings.
Advisor correspondence triage
A single thread view across counsel, accountants, and managers.
Household operations
Budgets, schedules, vendor records, and property files.
Governance and decision history
A searchable record of what the family has previously decided.
The system supports the principal and the office. It does not replace counsel, accountants, or investment advisors — it makes the staff that coordinates them dramatically more effective.
Principles
These constraints are baked into every office deployment. Privacy is not negotiable — it is the architecture.
The family controls the perimeter.
Models, documents, and chat live on hardware or in a single-tenant environment under your direct control. No shared inference. No vendor staff with access.
Principal-only documents stay principal-only.
Access is scoped by role, family branch, and trust. A family member can ask about their own documents without seeing the principal’s strategic files.
Every answer is grounded in a document.
Outputs cite the page, the file, the version. Nothing fabricated. Where the model has no source, it says so. Material decisions are always verifiable.
No vendor lock-in on the inference path.
The system runs on open-source models you can swap. If a vendor changes their terms, your office does not have to rearrange.
Example deployment
A reference design for a single-family office. The same architecture extends to multi-family offices with per-family scoping, and to private wealth operations that need the rigor of a family office without the headcount.
A private AI computer deployed in a single-tenant environment under the office’s direct control. Documents are ingested behind the family’s perimeter. The principal, the chief of staff, and named advisors have scoped access. Outside vendors never see the archive.
Document types included
Core workflows
Security model
Have a similar workflow?
Book a call →Reference deployment · single-family office
Who this is for
This is for principals and chief-of-staff equivalents whose job is to keep the family’s information inside the family. Where a single misplaced document is not a vendor’s problem — it is a family problem.
Strong fits
Common patterns
Especially good fit if you hear
In their words
“I spend the week before a family meeting in a stack of PDFs. With this, the briefing pack is drafted by Sunday and I am editing, not assembling.”
“There is no scenario where I put a trust document into a public AI tool. This is the version I can actually run.”
“It does not replace counsel or the CPA. It makes the staff coordinating them three times more effective. That is exactly what we needed.”
Deployment
The workspace is built so the family — not a vendor — controls the perimeter. We deploy on your hardware, in your private cloud, or in a single-tenant environment your office already trusts.
On-premise rig
A dedicated AI workstation in the office or at a property. Models run locally.
Single-tenant private cloud
For offices that prefer a managed environment without shared inference.
Open-source models
Llama, Qwen, Mistral, and Whisper. No third-party API calls on the inference path.
Scoped access
Principal-only documents, family-branch scoping, and named-advisor read access.
Document ingestion
Years of estate, investment, tax, and household records indexed for search and retrieval.
Citation-grounded outputs
Every answer ties back to a document and a page. Nothing fabricated.
The point is to give the office the leverage of a much larger operation — without ever turning the family’s private archive into someone else’s data.
What happens on the call
We talk through how the office actually runs — where the chief of staff’s time goes, where the principal is the bottleneck, and which workflows would benefit most from private AI.
We will map
You leave with a rough architecture, a priority workflow map, and an honest view of whether this is worth building for the office.
Project packages
Most offices either start with a focused audit or jump to a foundational build keyed to the next family meeting or tax cycle.
Workflow audit
For offices unsure where AI helps
A short engagement that maps how the office actually runs. You leave with a buildable architecture, a priority list, and an honest go/no-go.
Foundational office build
For offices ready to deploy
A working AI computer — archive search, estate and trust Q&A, investment document summaries, and meeting briefing preparation. Deployed in a single-tenant environment under your control.
Full build + automation
For multi-phase deployments
Foundational build plus household operations, advisor correspondence triage, governance and decision-history search, and ongoing support across tax and family-meeting cycles.
Family-office workspace builds commonly land in the tens of thousands depending on archive size, deployment model, and the level of integration with outside advisors.
Security
The deployment is designed so the family — not a vendor — controls every layer of the stack. No third-party AI APIs on the inference path. No vendor staff with access to documents. A clear audit trail of everything the system does.
A family office that lets confidential information into a public AI tool is one disclosure away from a problem. This is the version that does not put you there.
Who builds this
We work with a small number of offices each year. Hands-on, founder-led, end-to-end. The same rigor we have brought to large private deployments, scaled to fit a single-family office.
Brian Joseff
Co-founder
Designs the architecture and the deployment perimeter for each office. Background in private-cloud AI for sensitive environments.
Chandler Supple
Co-founder
Leads delivery and ongoing support with each office. Background in operational AI for document-heavy, advisor-coordinated work.
FAQ
No. The standard deployment isolates the archive inside your perimeter. We support the software layer through a secure channel that does not grant us access to your documents, prompts, or outputs. Where you need maintenance access, it is time-boxed, logged, and granted by your office on a per-incident basis.
For families and chiefs of staff who need the leverage of a much larger operation — and zero exposure of private information.