The questions we actually get.

No. Raptor is a technology provider. We do not hold client money, provide investment advice, manage portfolios, provide liquidity, or act as counterparty to any trade. Our clients are licensed firms who operate the technology under their own permissions.

No. If you are an individual trader you would reach Raptor through a brokerage that licenses it. We license to firms, not to retail clients.

A trading terminal, a client portal, a CRM, a back office, a risk engine, IB and affiliate management, an API across four transports, and the intelligence layer. Modules are licensed separately.

No, and this is worth being precise about: Raptor connects to the venues and providers you have contracted with directly, routes according to rules you set, and measures how each one performs. It does not provide, price or guarantee liquidity.

No. An EA runs a fixed rule set until you stop it. EMIL re-scores its own inputs against what actually happened, states conditions as a reading rather than a prediction, logs every event in readable language including its refusals, and is bounded by limits enforced outside itself in the order path.

Only when armed, only in the markets you selected, and only with the permissions you granted individually — open, close, modify, hedge. Granting close without open is common: it may reduce risk but never create it.

Arming requires reading an authorisation text and typing a confirmation word. Every action it proposes passes the same pre-trade checks as an order you place yourself, and nothing it concludes can widen its own limits. Disarm is one always-visible button, no confirmation, and Escape works too.

We will not claim that, and you should be wary of anyone who does. Adaptation reduces the chance of acting on a relationship that has stopped holding. It does not make a system right more often, and it does not remove market risk.

You can declare a portion of the account untouchable, set a profit floor that ratchets upward as gains are banked and never downward, a drawdown guard measured from the high-water mark, and a daily loss budget. All four are enforced in the order path, outside the intelligence layer.

It depends what you configured, and the behaviour is explicit rather than implied: stop opening, stop acting entirely while continuing to observe, or disarm automation and hand control back. We default the drawdown guard to disarm rather than liquidate, because forced selling at the worst price is itself a risk event.

A refused order always names which check refused it, the value the order would have produced, and the ceiling it would have crossed. We never silently reduce an order to fit.

Yes. REST for state, WebSocket for streams, FIX for institutional flow, webhooks for events. The interface itself uses the same endpoints we publish, which is the only durable way to keep an API current.

Generally no. Payments, verification, communications and market data sit behind adapter interfaces, so a provider is a configuration choice. Send us your vendor list and we will tell you plainly what exists and what would be development work.

Managed cloud infrastructure we operate, or your own environment. That choice materially changes the commercial shape.

We publish no certification badges, and we will not imply one we do not hold. The security page lists the controls that are actually implemented in plain language. Send your due-diligence questionnaire and we answer what we can evidence, and say so where we cannot.

There is no published price list, because a two-person prop desk and a multi-entity brokerage are not the same product. What drives the number is documented: modules, account count, instruments and venues, entities and brands, who operates the infrastructure, and what you migrate.

It depends almost entirely on migration and on your own onboarding and compliance requirements rather than on the platform. We scope it explicitly instead of quoting a number that assumes nothing goes wrong.

Yes, and we would prefer it. A demo is configured against your instruments, your session hours and your risk policy — including deliberately breaching one of your own limits so you can read the refusal.

Reference conversations are arranged case by case with the client’s agreement. We do not publish client names or figures without written permission, which is also how we would treat yours.