Workflow 01
Workflow API
Risks, controls, incidents, access requests, change requests, attestations — one model, one API, one surface for your second-line team.
Private alpha · Engineered in Zürich · Mapped to FINMA Circular 2023/1 controls
Governance, risk and compliance for FINMA-supervised firms.
One platform for risks, incidents, access requests, change workflows and audit evidence.
Unify the work your second-line team already does into a single audit trail. AI stays suggestion-only. Humans approve critical decisions. Engineered in Zürich. Hosted in Switzerland.
- Engineered in Zürich
- Product, team and hosting in Switzerland
- Private alpha
- NDA-based evaluation only
- FINMA 2023/1
- Mapped control catalogue
One real workflow
From access request to audit trail — in one system.
What a second-line team usually spreads across a ticketing tool, a compliance sheet, and a mailbox, Ordinis runs as a single traceable flow.
- 01Initiate
Access request filed
An operator requests a permission or scope change. Context, business justification, and linked entities are captured at the source.
- 02Validate
SoD + policy check
Segregation-of-duties validators run before anything is approved. Conflicts block the request with a traceable reason.
- 03Assist
AI suggestion (not decision)
Ordinis suggests routing, flags adjacent risk, and surfaces prior approvals. The suggestion is an audit row — it never mutates state.
- 04Approve
Human sign-off
The accountable approver accepts or rejects with a reason. Dual-control and escalation rules apply where the policy requires them.
- 05Record
Evidence + audit chain
Decision, supporting evidence (hash-pinned), and downstream notifications land in an append-only audit trail that auditors can read directly.
The same pattern runs incidents, change requests, policy attestations, and risk recalculation proposals. Event propagation, evidence handling, and the audit chain are shared — no parallel ledgers.
Three flow maps · interactive
Three views of how Ordinis works.
Switch tabs to follow risk through its lifecycle, see how the AI pipeline stays inside your tenant, or trace a single human-in-loop approval. Click any node to read its role.
Business context informs risks. Controls and KRIs monitor them. Incidents loop back into re-scoring. Policies and attestations close the loop into one audit chain.
Swipe the diagram sideways to see the full flow.
Context Risk transformation Events Assurance
Context
Your business process map is the canvas
Risks, controls and KRIs all tie back to a process and a business unit. Change a process and every dependent control re-evaluates automatically — no manual cross-reference, no drift between the org chart and the risk register.
Context
Board structure shapes every approval
Committees, charters and delegation rules are first-class objects. Approval thresholds and review escalations derive from your governance map, not from hard-coded constants in a workflow tool.
Context
Outsourcing is part of the risk picture
Significant third parties (FINMA Circular 2018/3) sit alongside internal processes. Their risks, controls and incident exposure are reviewed on the same cadence as in-house functions.
Risk
One risk register, one source of truth
Risks live in one register, not in three SaaS tools. Inherent and residual scores are visible side-by-side with the controls and assumptions that explain the delta between them.
Risk
Controls reduce risk — and are evidenced
Each control has an owner, a frequency and an evidence requirement. Re-review queues automatically when the underlying risk or the linked policy changes. Owners are not asked to chase the change manually.
Risk
Indicators surface emerging issues early
Threshold breaches generate signals — not noise. Suppression windows and dedup rules prevent the same condition from firing twice in a quiet hour, so reviewers see the breach once with full context.
Event
Incidents loop back into risk
Operational incidents and ICT events (FINMA Circular 2023/1) re-score affected risks and trigger lessons-learned attestations against the same audit chain. Nothing is filed and forgotten.
Assurance
Policies are versioned and signed
Policy changes propagate to every linked control and attestation. Old signatures stay immutable — you see the new signature happen, never overwrite the old, so the auditor can reconstruct who knew what when.
Assurance
Attestations are the human audit trail
Owners and reviewers sign off on policies, control effectiveness and remediation. Reminders, escalations and expirations are workflow primitives — not an email chain that lives outside the platform.
Assurance
Everything lands in one audit chain
Every change, approval, attestation and incident is recorded as an append-only row with integrity hashes. The Prüfgesellschaft does not have to reconstruct the flow from four different ledgers.
Click any node to read its role. Tab between maps for different views of the same platform.
These visualisations simplify the live engine for clarity. Full rule definitions, model inputs and audit-chain contracts are shared under NDA during the evaluation window.
How it works
Six capabilities. One audit trail.
Ordinis is built so the second-line team can do its actual work — not stitch together four disconnected SaaS ledgers with a spreadsheet in between.
Audit 02
Append-only audit chain
Every state change is recorded with integrity hashes. Tamper-evident by design. Auditors can read the trail directly, without a data-science round-trip.
Propagation 03
Event propagation
When a policy or control changes, dependent attestations reopen, notifications fire, and related reviews queue themselves — declaratively, with silence windows and dead-letter handling.
Evidence 04
Evidence store
Attach files to any decision. Artefacts are hash-pinned on write. Download is through short-lived signed URLs. Primary storage lives in Zürich.
AI 05
Explainable AI suggestions
Classifications and routing arrive as suggestions with context. Humans accept, dismiss, or override. The model never mutates regulated state on its own.
Access 06
Role + SoD enforcement
Fine-grained permissions and segregation-of-duties validators run before approvals. Drift is detected and surfaced on the permission side, not found during the next audit.
Who it's for
Built for the Swiss regulated perimeter — two wedges first.
Ordinis is tenant-native from the first migration and FINMA-shaped where it counts. During the alpha we're focused on two buyer types.
FINMA-supervised FinTechs
Licensed under FINMA Art. 1b (FinTech licence) or Art. 3 (bank licence). Map your risk register, controls, and incident pipeline to FINMA Circular 2023/1 controls on day one — with the workflow engine the regulator expects to see.
- ICT operational risk register linked to controls
- Third-party outsourcing controls + evidence handling
- Incident classification with AI-suggested banner (human approves)
Asset & wealth managers
FINMA-supervised asset managers and trustees. Run the full audit cycle — Prüfstrategie, Prüfgebiete, Prüfberichte — with Swiss-formatted reports and the audit chain your Prüfgesellschaft will read directly.
- Prüfgebiet templates + coverage heatmap
- SoD-enforced finding escalation + carry-forward
- Annual Prüfbericht artefacts with integrity hashes
Insurance intermediaries and other regulated segments are on the roadmap but are not in the current alpha. We'd rather go deep on two wedges than shallow on five.
Fit check
Who Ordinis is not for.
If one of these describes you, a shorter conversation saves both sides time.
Generic SME compliance
If FINMA, FADP, Swiss audit workflows, and second-line governance aren't central to your operating model, Ordinis is over-specified for you.
Checklist-only GRC
Ordinis is a workflow engine with an audit trail underneath. If you need a policy library to tick boxes against, there are lighter tools.
Spreadsheet replacement without workflow
We don't ship a risk register you fill in manually. Risks connect to controls, incidents, propagation rules, and approvals — or they aren't worth tracking.
What second-line teams get
Core capabilities, plainly described.
What's shipped today. Where a capability is partial or on the roadmap, we say so.
AI-assisted, never AI-decided
Every classification or routing suggestion is written as an audit row with context. Your team accepts, dismisses, or overrides. The model does not mutate regulated state on its own.
Event-driven propagation
Declarative rules express how a change to one entity ripples — reopening attestations, queueing reviews, notifying owners. Dedup gates, silence windows, and dead-letter handling are first-class.
Hosted in Zürich
Primary application and data live in Switzerland. Evidence and generated reports are stored in the same region. Secondary services (observability, some AI) are disclosed per-service, not hidden.
FINMA-shaped entities
Circular 2023/1 controls, BankV Art. 7 outsourcing register, GwG audit points, and Prüfstrategie/-bericht flows are first-class entities — not spreadsheet bolt-ons you paste into the product.
Role + SoD enforcement
Fine-grained permissions ship through a drift-detection pipeline. Segregation-of-duties validators run before state mutation on access requests, change requests, and risk decisions.
Tamper-evident audit chain
Append-only audit log with integrity hashes. Evidence attachments are SHA-256-pinned on write. Auditor-ready exports without a data-scientist in the loop.
Compliance posture
Regulator-shaped. Auditor-defensible.
Each claim below maps to a demonstrable control. Certifications still in progress are listed as such — not implied.
- FINMA Circular 2023/1
- Operational risk + resilience — mapped in the control catalogue
- FADP (Swiss DPA)
- Aligned by design; DPIA artefacts supported
- GDPR aware
- EEA data-subject workflows available
- Engineered in Zürich
- Product and engineering in Switzerland
- Hosted in Switzerland
- Application, data, and evidence in ch-dk-2
- Tamper-evident audit chain
- Append-only log with integrity hashes
- Explainable AI suggestions
- Context + provenance on every AI row
On the roadmap, not in place: SOC 2 Type II certification, multi-region tenant-configurable residency, independent penetration test report.
Fully shared under NDA: specific FINMA control-to-entity mappings, audit trail schema, evidence integrity model, and AI insertion rubric.
Current alpha scope
Where Ordinis is today, plainly.
We're in private alpha. That changes how you should read everything on this page.
- Private alpha
Closed cohort. Access is granted case-by-case after a short conversation.
- NDA-based evaluation
All access operates under a bilateral non-disclosure agreement. No public self-serve signup.
- Core workflows available
Risks, controls, incidents, access requests, change requests, policy attestations, audit evidence.
- No public pricing yet
Pricing is set after the alpha cohort closes and feedback is incorporated.
- Roadmap is transparent
Missing capabilities (SOC 2 Type II, multi-region residency, independent pen-test report) are disclosed up front — not hidden behind marketing copy.
FAQ
Things people actually ask.
What stage is Ordinis at?
A private alpha instance runs in Zürich. A small number of design partners are evaluating the first end-to-end workflows under NDA. Public launch, pricing, and broader availability follow once that cohort's feedback is incorporated.
Where is the data stored?
Primary application and data live in Switzerland (Zürich, ch-dk-2). Evidence attachments and generated reports use object storage in the same region. Secondary services (error telemetry via Sentry EU/Frankfurt, some AI APIs) are disclosed per-service — we don't claim blanket Swiss-only hosting.
How does the AI layer work — and is it decision-making?
No. AI is suggestion-only. Classifications, routing, and risk-score attributions arrive as audit rows with context. A human with the right role accepts, dismisses, or overrides. The model does not mutate regulated state on its own — that's an architectural commitment, not a toggle.
How is Ordinis different from a GRC spreadsheet vendor?
Events propagate. When a policy or control changes, dependent attestations reopen, notifications fire, and related reviews queue themselves — declaratively, not manually. Every action is idempotent, audited, and provable against the same data model a regulator would read.
Can I bring my own compliance framework?
Yes. FINMA Circular 2023/1, GwG audit points, BankV Art. 7, and Swiss audit workflows are seeded. New frameworks are configured through the regulation catalogue plus the propagation-rule registry — not hardcoded.
What is not in the current alpha yet?
SOC 2 Type II certification is in progress. Multi-region tenant-configurable residency and an independent penetration test report are roadmap items. Insurance-intermediary workflows are not in the current alpha. ISO/IEC 27001:2022 is already in place — Ordinis is built under Finray's certified information security management system.
Who is behind Ordinis?
Ordinis is built by Finray Technologies Limited, a Cyprus-headquartered financial-technology firm. The core team is based in Zürich and across the EU, combining senior FinTech, compliance, and engineering leadership.
How do I get alpha access?
Request access below. We review each inbound to make sure the pairing is useful for both sides — current cohort is limited to FINMA-supervised FinTechs and asset managers.
Powered by Finray
One platform in the Finray ecosystem.
Ordinis is the flagship governance, risk and compliance product of Finray Technologies — a financial-technology infrastructure provider. While Ordinis is your operational brain for regulatory work, it runs under Finray's ISO/IEC 27001:2022 certified information security management system.
Ordinis is the governance, risk and compliance layer. Other products in the Finray ecosystem extend the same integrity posture to settlement, treasury and reporting workloads.
Shared with every Finray product
- Supervisory-aligned Built to the controls auditors actually read — FINMA Circular 2023/1, FADP and the Swiss audit framework are first-class concerns, not retro-fit mappings.
- Explainability first Model outputs arrive with drivers and regulation references attached. AI stays suggestion-only across the portfolio. Human approval is not a toggle.
- Integrity by design Append-only audit chains with integrity hashes. Primary application and data hosted in Switzerland. Per-service hosting is disclosed, not hidden.
Finray Technologies is the operating entity behind Ordinis. The same engineering posture extends across the rest of the portfolio.
Private alpha · Design partners under NDA
Become a design partner.
We're working with a small number of FINMA-supervised FinTechs and asset managers under bilateral NDA. If Ordinis fits what your second-line team actually does, we'll provision an isolated tenant and pair for the first two weeks.
Inbound messages are processed under our privacy policy. We won't add you to a marketing list. No public self-serve signup.