The Core mandates and the live systems, delivered as a sleeve your desk allocates to, monitors, and can switch off at will. Capital stays in your name, positions stay visible, and nothing is pooled.
The first problem is what it costs. Expense ratios, management fees, and performance fees that take a fifth of a good year compound against a family office in the one direction that matters. They are charged whether or not anyone earned them: on the fund that bought the top because inflows arrived that morning, and on the manager who has not looked at the position since. The fee buys the mandate to deploy. It does not buy the judgement to enter well.
Over a decade that drag is not a line item. It is a meaningful share of the wealth the family actually built.
The second problem is that the good sleeves are shut. Capacity is full. The fund stopped taking outside capital two vintages ago. The allocation you want is open only to whoever was already inside when it was open, and the answer you get is a waiting list rather than a term sheet. Being right about a strategy and being able to access it have never been the same thing.
An AI desk carries neither constraint. It runs the same way at ten million as at two hundred, it does not close a vintage, and it arrives in an account your desk already controls rather than in a vehicle you have to be admitted to.
Algo Alpha was founded by two people who still run it, supported by engineering and quantitative research teams drawn from institutional and defence programmes.
Co-founder. Sets the research agenda, the risk framework every system trades under, and the standard that no strategy is released until it has run on our own capital first.
Co-founder. Runs the institutional relationships, from first conversation through mandate selection, account structure, and the reporting a committee actually needs.
Builds the execution layer, the risk controls, and the infrastructure the mandates run on. Alumni of programmes where a failed system was not a support ticket.
Builds the valuation frameworks and the analysis that decides when the evidence is there. Research first, engineering second.
What protects an allocator is rarely a promise. It is the list of things the manager has no ability to do. Ours is short, and it is enforced by the account structure rather than by our own good conduct.
Have AI managed portfolios with proven strategies engineered for performance, directly in your own account.