Book a demo

Private wealth · AI distribution live

Financial products, distributed by AI.

New clients through the assistants they already use. New products through rails the institution does not have.

Book a demoHow it works

Distributed inside

  • ChatGPT
  • Claude
  • Gemini

The institution's product. Its client. The client's assistant.

Client assistant

Live

I've got 250k sitting idle. What should I do with it?

Your mandate allows a USDC yield strategy. Here is the one your private bank distributes:

Odyssey · EUR Yield Vault

6.4%

Net APY

Daily liquidity

Mandate-checked

Your banker is notified

01 — The problem

The demand for financial products has already moved into AI assistants. The products can't follow.

$84tn

The generational transfer

Passing to a generation that has never known banking without a smartphone — $22tn of it to Gen X and Millennials within ten years.

81%

Switching intent

of next-gen HNWIs plan to switch away from their parents' wealth management firm within one to two years of inheritance.

47%

AI-first

of wealthy investors aged 21–45 already use AI to research companies or markets. They don't book a meeting — they ask an assistant.

The next generation does not book a meeting. It opens an assistant.

Capgemini World Wealth Report, next-gen HNWI switching intent · Bank of America Private Bank, 2026 Study of Wealthy Americans ($3m+ investable assets), investors aged 21–45 · $84tn / $22tn intergenerational transfer, ten-year horizon.

02 — Solution

Odyssey packages financial products for AI — and runs the rails they execute on.

Client side

  • Client asks in the assistant
  • Portfolio-aware reasoning
  • One-click execution

Odyssey layer

Packaging

Raw primitives become finished products with live pricing.

Mandate engine

Suitability checked before anything is shown or signed.

Execution rails

Custody-level enforcement, settlement, reporting.

Institution side

  • Products stay the institution's
  • Client intent data, live
  • Nothing to migrate

One layer. Three surfaces. Nothing for the institution to migrate.

03 — Signature use case

From 250k sitting idle to an executed, mandate-checked position.

    01

    The client asks

    “I have 250k sitting idle. What should I do with it?” Asked in the assistant, not in a meeting booked three weeks out.

    02

    The agent proposes

    A product from the institution shelf, checked against this client mandate before it is ever shown: net APY, liquidity, risk score.

    03

    The client acts

    One click builds the transaction he signs himself. Or he does nothing — which is a signal too.

    04

    The institution sees

    Full-context intent on the console: what was explored, what was funded, what was left on the table.

04 — The institution view

The institution sees the intent — and the mandate that shaped it.

01

Valentin opens a Claude conversation

“I have $250k idle — what can I do?” The assistant reads his live portfolio, not a brochure.

02

The mandate layer filters the shelf

3 products eligible, 3 hidden. A $50k PEPE request is refused before it reaches the banker.

03

The agent proposes, the client signs

$100k → MoonETH and $150k → MoonUSDC, mandate-checked, awaiting client signature.

04

The banker gets signal, not noise

2 proposals pending 31 min · 1 out-of-mandate attempt. He knows who to call, and why.

One mandate per client: the banker edits the limits, the agent stays inside them — every conversation lands as a signal, not a ticket.

Institution console — live demo

Live
Mandate per client, pending proposals and refused out-of-mandate attempts — mirrored on-chain via Zodiac Roles.

05 — Traction & proof

We already operate what we sell.

$3m

AUM · +100% in 12 months

Live tokenised funds on Ethereum mainnet, via Lagoon. Verifiable NAV, on-chain.

~$100k

Revenue generated

Earned by Odyssey Digital AM on our own book, before a single line was sold.

5

Institutions in dialogue

Private banks, EAMs and allocators, live on the product shelf.

Live

Working demo

A custom Claude connector distributing Odyssey products, execution button included.

Client pull, before the product: existing clients were already asking to reach their portfolios through Claude — before we built the channel. Every number here is verifiable on-chain or in the accounts.

06 — Who pays

Net-new revenue, zero core-banking integration.

New products on new rails: nothing to migrate, nothing cannibalised. Revenue on assets the institution cannot see today, from clients it is otherwise about to lose.

  • Asset managers
  • Private banks
  • Wealth & brokerage platforms
  • EAMs & family offices
  • Neobanks & fintech distributors

Precedent, not experiment

Institutions buy this. They do not build it.

Three peers have already made the same call, and none of them started from a blank page. A conservative buyer says yes quickly when nothing it owns has to change.

07 — Team

Two founders. Institutional operators who run the book themselves.

Ex-J.P. Morgan

Yacine Boughazi

Co-founder & CEO

Digital assets M&A. Owns the bank channel, the raise and the regulatory perimeter.

Bank channelFundraiseRegulatory perimeter

Ex-BCG

Théo Viggiano

Co-founder & CTO

École Polytechnique, HEC. Built the vault architecture, the multi-Safe setup and OdysseyRisk.

Risk engineExecution railsVault architecture

Design partners · UAE & Switzerland

Put your shelf inside the assistant your clients already use.

Two design partners first, then the shelf every institution plugs into. One product reaches every institution. One institution reaches every product.

Become a design partnerExplore the platform