Three seats exist to make the firm argue with itself before it commits. None of them decides anything. The general partner makes every call, and the record of those calls is kept and graded.
The bear case is the step a firm skips when it likes a deal, and the deal it likes is the one that most needs one. The Skeptic writes the case against, and it is kept away from the Analyst's work on purpose: an analyst asked to argue against its own read is grading its own homework.
It is allowed to fail, and that is its most valuable output. A skeptic who always finds something is a skeptic nobody weighs. Tried to kill this and could not, and here is what was tried, is the single most useful sentence available before an investment.
Anyone can produce a confident view. The Forecaster makes a call on the outcome and is measured against what actually happened, so the firm can see whether its machine is any good rather than assuming it.
A read can also say watch, which is the answer a great many deals deserve and most systems have no way to record.
Every decision is logged as it is taken and returned to the moment there is a verdict to compare against. A firm that does not grade its own decisions repeats them, and a record reconstructed a year later is a record written from memory.
None of this is a committee. These seats do not vote and do not decide. They argue, forecast and record, and then you make the call.
A committee decides. These seats do not. The Skeptic argues, the Forecaster predicts and is scored, the Historian records, and then the general partner makes the call and owns it.
That is not modesty about what the software can do. It is the shape of the product: a firm that hands the decision to a system has swapped one kind of risk for another it cannot explain to its LPs, and no part of this is built to accept that trade.
What it does change is the quality of the argument you are having with yourself. Most solo GPs decide alone, and deciding alone means the strongest version of the objection is whichever one you happened to think of. Here it is written down before you commit, by something that never saw the case for.
The case against, scored forecasts, and a record of decisions: Skeptic, Forecaster, Historian. A firm of 20: you and nineteen AI employees. See the whole firm and join the waitlist.