Canadian-only processing
All inference for Oshan’s Augure-hosted models happens in Canada. Generation, embedding, reranking and document processing run on Canadian infrastructure.
Find what your company already knows with Oshan.
Search your organization’s documents in English and French. Inspect the sources behind an answer, with access and deployment configured for your team.
From a question to a source you can check.
Read the evidence beside the answer.
Open a citation to read the source document. Validate answer quality, unsupported questions and conflicting versions against your own documents during the pilot.
Starting at $50 CAD per user / month
Hosted inference is included within your agreed allowance. Your pilot proposal sets the token pool, seats, onboarding costs and overage policy. Bring-your-own-model usage follows your provider’s terms.
How much time would it need to save?
Set your assumptions. Every purchased seat contributes to licence cost; only active users contribute to the value of time saved.
Estimated annual net time value
$13,200
Break-even: 5.6 minutes per active user, per day.
- Purchased seats
- 50
- Active users
- 30
- Annual licence
- $30,000
- Annual time value
- $43,200
Assumes 240 working days, in CAD, before tax and onboarding. Time value is not guaranteed cash savings. These inputs are not measured Oshan performance.
A pilot built around your systems.
- 01
Scope
Choose the source, pilot group, deployment policy and representative questions.
- 02
Connect and validate
Index the approved corpus. Test both languages, citations and access boundaries.
- 03
Roll out deliberately
Set seats and budgets, rehearse recovery, then agree on wider access.
Choose where storage and processing happen.
For an Augure-hosted deployment, customer storage and model inference are both in Canada. Bring-your-own-model and self-hosted setups are documented in your deployment plan.
Bring your own models
Use the models your organization approves. Confirm their processing location, retention terms and usage charges as part of the deployment.
Your infrastructure
Operate Oshan in your own environment. A disconnected deployment requires local models, preloaded artifacts and an offline operational test.
Before you start
Read about securityWhat does the starting seat price include?
Pricing starts at $50 CAD per user per month. Hosted inference is included within the allowance agreed in your proposal. Seats, token pools, onboarding charges and any overage terms are confirmed before purchase.
Does access automatically follow my source system?
Not in the current pilot release. Access is assigned through administrator-managed users and groups for an approved corpus. Automatic source-permission synchronization is unavailable until its indexing and revocation lifecycle has been validated.
Can we keep processing in Canada?
Yes. With Oshan’s Augure-hosted models, generation, embedding, reranking and document processing all run in Canada, and customer storage is in Canada too.
How is bilingual quality tested?
The pilot evaluates English-to-English, French-to-French, French-to-English and English-to-French retrieval, along with citations, absent answers and conflicting documents. Results are checked on representative documents from your organization.
Where is our data stored?
Augure-hosted customer storage is in Canada. For customer-operated deployments, your infrastructure determines storage location. If you bring your own models, their processing location and terms are documented in your deployment plan.
Let’s test Oshan against your real questions.
A demo leads to a scoped pilot proposal: sources, access model, allowance, deployment and acceptance criteria. No self-serve signup required.