One database for the whole hardware business.
A leasing tool, an ITSM, a resale spreadsheet and a disposal vendor’s portal rarely agree on what one fleet actually cost you. Cevoriq keeps one record per device, from the purchase order to the wipe, so that number is a query instead of a week of reconciliation.
Built for DaaS providers, field-service teams, hardware resellers and ITAD shops.
What Cevoriq covers
Most customers arrive for one of these and switch on the rest later. Modules are licensed separately, so a workspace only shows what it actually uses.
Device as a Service
Fleets you manage for customers
Each customer fleet sits in its own sub-account under its own contract. Devices carry a health score built from age, repair history, warranty status and incident rate, which turns refresh planning into a queue you approve rather than a spreadsheet you maintain.
End-of-term is priced, not guessed: residual value runs down a depreciation curve that a disposition partner can also underwrite for a buyout, so “buy it back or ship it back” is a number before the lease actually ends.
- Customer sub-accounts
- Health scoring
- Usage metering
- End-of-term disposition
ITAD
Retiring assets with the paperwork intact
Every unit is serialized and graded at intake, then priced against valuations you have approved before it routes to resale or disposal.
Data sanitization is a field on the asset, not a checkbox on a form — set it once to the standard a customer requires (a NIST 800-88 purge, or their own policy), and the wipe record that satisfies it stays attached to that serial number through to its terminal state.
- Intake grading
- Disposition pricing
- Chain of custody
Field Service
On-site work, by your own technicians or by partners
Your own technicians work from a scheduling board and a field app that keeps working offline, recording photos, checklist answers and signatures as proof of service. Partner work is priced from each partner’s rate cards and country coverage, dispatched by hand or auto-assigned within a spend ceiling you set, and taken in a partner portal or in the subcontractor’s own FSM tool. Tickets sync both ways with ServiceNow.
Scheduling board · Offline field app · Proof and signatures · Maintenance plans · Coverage check before dispatch · Escalation · Capacity and crews · Vehicles and van stock · Fleet data feeds · Partner portal · Partner compliance documents · Partner scorecards · Rate-card pricing · AI auto-assignment
Resale
Selling hardware without holding stock
Customers order from your catalog, you source from an OEM or a distributor, and the goods ship straight to the customer. The ledger records the pass-through and keeps your margin on its own line.
Dropship · Pass-through ledger
Where the AI sits
Drop an MSA in and it pulls out the parties, contract value, term length, SLA commitments and termination clauses, notes whether the contract renews automatically, and quotes its renewal, liability and payment terms as written. It then scores the risk, and when a reviewer corrects a score, later contracts like it are scored with that correction in view.
We don’t hide the logic. Every recommendation shows its working: the six factors behind a device’s health score, the rate cards behind a buyback estimate. If an automated action makes you nervous, set the confidence it has to clear. Anything under the line goes to a human queue. Any rule can be switched off, and the decisions it already made stay on the record.
Automated dispatch runs against a ceiling you set per customer — a spend amount or a job count, per calendar month. When a ticket would take you past it the work does not vanish; it lands in your queue for a human, who is not capped.
Point a workspace at your own Anthropic, Google, OpenAI, Azure or AWS Bedrock key and the prompts leave on your subscription, not ours. Workspaces without keys use ours, metered per tenant.
Device health, scored
Weighted factors: age, repair history, warranty and incident rate. DEX telemetry and performance trend are scored when a device carries telemetry; no DEX connector is available yet, and a factor with no data is left out rather than counted as healthy.
| Score | Band | Meaning |
|---|---|---|
| 0 – 30 | Keep | Healthy. No action. |
| 31 – 55 | Monitor | Watch for early signs. |
| 56 – 80 | Plan refresh | Schedule a replacement. |
| 81 – 100 | Urgent refresh | Replace now. |
Refresh candidates get a second pass that recommends reassign, repair, resell or dispose.
Connections
Customers and vendors can hold their own authenticated connection into your workspace, each scoped to exactly what it may see. Orders arrive without re-keying, and your catalog publishes outward so buyers only order what you can actually supply.
On the tooling side, incidents sync both ways with ServiceNow. Supplier catalogs, vehicle data and a subcontractor’s own field-service tool connect through a mapped REST connection you configure; vehicle data can also come in as a CSV upload. The presets below are written from each provider’s public API documentation and have not yet been run against a live account.
- ServiceNow
- OEM and distributor catalog APIs
- Samsara (preset)
- Geotab (preset)
- ServiceNow Field Service Management (preset)
Planned, not yet available:
- Jira Service Management
- Zoho Desk
- Freshservice
- Zendesk
- Nexthink
- ControlUp
- Lakeside SysTrack
- Microsoft Intune
- Jamf Pro
- FedEx
- UPS
- DHL
Security
Cevoriq runs fleets that belong to your customers, so isolation between organizations is the part we are strictest about.
- Tenant isolation
- Every query runs as a restricted database role under Postgres row-level security, scoped by a signed session token the client cannot forge. A suite of 28 adversarial queries — cross-tenant reads, tenant re-pointing, forged seals — runs against a real Postgres in continuous integration and fails the build, not warns, if any one of them returns a row it should not.
- What we found when we attacked it
- A twenty-phase internal review ran in September 2026 against a disposable tenant in production, not a staging copy — live exploitation attempts, not only source review. An unscoped query aimed through our own reporting feature at a table holding other tenants' live data returned only the caller's own rows. Six items needed attention and five are fixed; the sixth is a product-policy question, not a defect. The tenant was then destroyed and every credential minted for the review confirmed dead. An independent third-party test is the next step, and has not happened yet.
- Your AI boundary
- On bring-your-own-AI, prompts and data travel on your own provider subscription. Nothing is pooled onto ours.
- Cross-tenant reporting
- A report can span organizations only where the owning tenant has recorded a typed, server-verified approval.
- Deletes and legal holds
- A delete is a marked intent with a declared purge date and a recovery window, so evidence survives an accident. Records under legal hold can be retrieved from archive.
- Residency and governance
- Declarative data-residency controls, per-tenant AI processing consent, GDPR data-subject erasure, and a full audit trail. Choose where an organization's data lives at creation: US East (Virginia) or EU Central (Frankfurt). It is stored in that region and no other.
- Access control
- Owner, admin, manager, viewer and technician roles, where a technician sees only their own jobs, plus per-member module scoping. Both are checked on the server, not only hidden in the navigation.
Platform
What a general ITAM tool doesn’t do
Partner pricing that survives contact with a deal
Field-service partners are priced two different ways and most tools model only one. A standing price book collects indicative rates across countries, services and response tiers; approved lines become the catalog that dispatch actually prices against.
A deal round is the other kind. It belongs to one customer, closes on a hard deadline, and the quotes are binding. Partners answer through a link with no login and see only their own numbers. Approving a deal quote deliberately does not touch the standing catalog — award it and the winner’s prices become customer-scoped overrides, so nobody else gets repriced by your negotiation.
The award produces a routed engagement: exclusive, ranked with failover, or a broadcast pool. If the rank-one partner leaves an offer unanswered, or the SLA clock closes in while it is still unaccepted, the work moves to rank two on its own and the reroute is written to the audit log.
SLA clocks that follow the work, not the head office
An SLA is only meaningful if it runs on the right calendar. A ticket delivered to a site in Lisbon should breach against Portuguese business hours and Portuguese public holidays, not against whatever timezone the contract was signed in.
So the calendar resolves from the delivery site’s country first, then the contract, then the customer, then the tenant default. Response, resolution and attainment are measured on that clock, with breach history kept per customer.
Also included
- Incidents linked to the device and contract involved, with the RMA lifecycle in both directions including excess-refund approvals.
- Ask a question in plain English and it runs as real, read-only SQL scoped to your tenant. Freeze a period into a snapshot and share it by an expiring, revocable link to someone with no login.
- Per-module licensing and metering, with individual members scoped to just the modules they should see.
- Warehouse, logistics and workshop for the operators who hold stock and repair it.
Getting started
- Import contracts, assets and incidents from CSV, or connect ServiceNow.
- Switch on the modules you run. Add your own AI keys if you want them.
- Work the review queue, and promote the decisions you keep approving into rules.
Ask for early access
We are onboarding our first customers now. Leave a work email and we will come back to you about a pilot.