Fasil

Fasil is AI decision-intelligence software for institutions that turns a contested question into a reviewable decision record. Rather than producing a recommendation, it governs the path from disagreement to decision - connecting the arguments made, the evidence behind them, the reasons each was evaluated as it was, and the outcome that followed - so the institution ends up with something it can defend and revisit later.

The operating layer is called Sijal. It examines a question through ordered arguments, evidence and reviewable evaluations, running the deliberation across six structured phases including rebuttal and refutation, with support and opposition positions kept distinct. Six AI agents with custom personas argue the positions, with real-time scoring and citations attached to each move. Sessions stay private inside the institution's own archive.

The output is a decision record showing its decisive weights and, importantly, what would have to change for the decision to change - which is what makes it auditable rather than just documented. Fasil operates in Arabic and English.

Access is by request rather than self-service signup: the site's call to action is "request institutional access," and no pricing is published. Aimed at institutions making high-stakes decisions that must survive later scrutiny - governance bodies, policy teams and committees rather than individual users.

Running structured institutional deliberation on contested questions and producing an auditable decision record of arguments, evidence, weights and outcomes.