Managed Investing
Hand the portfolio to a model.
See what the model costs.
Pick a risk profile and a model on Transaction Science AI runs the strategy. TradingOS executes the rebalances. Insights meters the energy and compute cost of every model run — including the runs where the model looks and decides not to trade — and writes it onto the rebalance record. The model's work is on the statement, not abstracted into a percentage.
Conservative
Capital preservation · short duration
Balanced
Broad market · income tilt
Growth
Equity-led · global
Aggressive
Concentrated · high beta
Three pillars on one record: the model is Transaction Science AI, the execution is TradingOS, the joules and the dollar cost are Insights.
A real model, not a glidepath
The strategy is run by a named model on Transaction Science AI — one of the non-transformer engines — with its id, version, and family recorded on every decision.
Executed by TradingOS
Rebalance trades route through the same matching engine, pre-trade risk, and surveillance as every other order on Settlement. No separate path for the managed book.
Metered by Insights
Every model run carries its joules, its compute cost, and its duration. You can see the lifetime energy the portfolio has consumed, the same way you see lifetime returns.
No-trade runs count too
When the model evaluates and decides to do nothing, that run still has a meter. You paid for the model to think; you see what thinking cost.
A management fee tells you what you paid. Not what it bought.
A managed portfolio charges a fee for the work. This one also shows the work: the rationale the model gave, the trades it placed, the joules it burned, and the compute it cost — on every run, in the record. It's possible because the model, the execution venue, and the meter are all Transaction Science platforms, reporting into one account. The fee is the easy number. The rest is usually invisible. Here it isn't.