Built to fit
We sell neither user tiers nor checkbox options: two companies of the same size have different sources, different volumes and different hosting constraints. The scope is defined with you, and the quote follows from it. Let's talk about it directly.
In practice
No online checkout and no plan to pick: the first step is a conversation.
Nothing is held back for a higher tier: every installation gets the whole platform. What is decided with you is its size and how it is run.
Connectors are not counted one by one: plug in as many sources as your scope requires. Cost follows the volumes processed, not the number of integrations.
Three parameters, nothing else. We go through them with you before quoting.
Number of users, indexed documents and questions asked each day. This is what determines the size of the installation.
Shared, managed dedicated instance, or installed on your own infrastructure. Running costs differ depending on who holds the infrastructure.
Taking over existing material, designing the first pipelines, training teams, level of assistance and guaranteed response times.
Language models are billed on actual usage. The platform counts tokens and embeddings separately, per job: you watch the spend build up, you do not discover it.
Four steps, with no middleman and no online checkout. You talk to the team that builds the product.
A few lines are enough: your business, what you are trying to connect, and which department comes first.
A one-hour conversation about your sources, your volumes, your hosting constraints and the teams who will use the platform.
A detailed scope and a firm amount, stating what is included and what stays usage-based. Nothing hidden in an appendix.
First scope connected, graph built, assistant opened to a first team. This pilot doubles as a proof of concept on your own data, and the installation then grows at your pace.
Because it would be wrong for almost everyone. Between a first shared scope for one department and a sovereign deployment on your own cluster, the real cost varies by an order of magnitude. Publishing a single figure would force us to set it high for all. We would rather quote after a conversation, and stand by it.
The sources to connect, a rough order of magnitude for the documents to index, how many people will use the platform, and your hosting constraint. If you do not know those figures yet, the scoping conversation exists precisely to establish them.
A few days for a first scope in shared mode: connect two or three sources, let ingestion build the graph, open the assistant to a first team. An installation on your own infrastructure needs more preparation on the network and cluster side.
By counting time, not licences. The time saved shows up on tasks your teams can already price: finding information scattered across several tools, reconstructing the history of a case, preparing a recurring summary. Take three of those use cases, record what they cost today in hours, and measure again after the pilot. That is the only honest way to work it out, because it rests on your activity rather than a market average. We publish no productivity figure for that reason: we do not have yours.
Model usage is billed on actual consumption, separately from the platform. You can use our provider or your own. Either way, consumption is counted per job and visible in the supervision consoles.
No. Your data feeds your graph and your answers, nothing else. When installed on your own infrastructure, it never leaves your network.
Your original documents are yours and stay where they are: Omniscient does not replace them, it links them. The graph and the logs can be exported in open formats.
Platform authentication is fully covered: accounts, predefined and custom roles, fine-grained access rules per resource and per action, password rotation and time-limited sessions. Connecting to your directory (single sign-on, LDAP, multi-factor authentication) is not a stock component you switch on: it is built specifically for your information system, as part of the installation. It is one of the things the scoping conversation is there to settle.
No, quite the opposite. Omniscient plugs into the CRM, shared folders and databases you already use. It does not replace your business tools, it makes their content searchable together.
A one-hour conversation is enough to produce a firm quote: your sources, your volumes, your hosting constraints.