Good questions come before the project.

Users, sites, criticality and internal IT determine scope, from small businesses to distributed organisations.

Replacement is not the default. We assess existing systems and modernise where security, lifecycle, cost or operations justify it.

Assessment, migration or modernisation can be commissioned separately. Ongoing support is agreed additionally if needed.

Services cover infrastructure, networking, security, cloud, managed operations, workplace and recovery, plus software, automation, consulting and technical leadership.

Describe the problem and goal. We ask about systems, risks, users, sites, timing and ownership to agree a suitable entry point.

Yes. A co-managed model divides tasks between your IT team and Endurion, with named owners for specialist work, projects and operations.

Remote and on-site work are agreed by task. Assessments, workshops, rollouts or critical changes may need an on-site visit.

The start date depends on scope and available capacity. Migrations and takeovers need access, owners, maintenance windows and an agreed start plan. Immediate availability is not generally guaranteed.

Support handles incidents and requests. Managed services additionally define recurring monitoring, patching, backup checks, security and reporting responsibilities.

Plans list response targets from eight business hours to prioritised Enterprise P1 handling. Your agreed SLA matrix is contractually applicable.

24/7 is a separately agreed option requiring explicit critical-service scope, on-call coverage, response targets and escalation.

Billing and term are defined in the proposal and contract, including notice periods and included services.

Published amounts and calculator results are non-binding. A quote requires review of scope, users, sites, licences, legacy systems, security and SLA.

Headcount explains only part of the effort. Licences, sites, legacy systems, migration and security can differ substantially between similar-sized companies.

Yes. The selected plan is carried into the proposal request. Add your situation and contact details there; user counts, sites and technical details can be provided as needed.

We support technical and operational preparation: inventory, ownership, controls, logs, backup, incident processes and evidence. Legal assessment remains separate.

The agreed incident process follows impact and urgency, covering containment, analysis, recovery, communication and follow-up improvement.

Access is checked by identity, device and context and limited to needed permissions, while legitimate work and exceptions remain supported.

Options include Azure, AWS, Google Cloud and private or hybrid environments, chosen by workload, requirements, skills, costs and exit options.

Full migration is not required. We assess each workload technically, economically and against regulatory needs; hybrid operation may fit.

Backup and recovery are planned together. RPO, RTO, protected copies and restores test whether backups support the required recovery.

Yes. Depending on scope, we assess or manage tenant and licences, Entra ID, Conditional Access, Intune, Autopilot, MDM, device standards and support.

Marks identify technologies and do not prove partnership. Any partnership or certification would need explicit, verifiable confirmation.

The directory links to Microsoft 365, Azure, AWS, Google Cloud, Cisco, Juniper, Fortinet, Palo Alto Networks, Veeam, VMware, Nutanix, Dell, NetApp, Proxmox, Red Hat, Sophos, Check Point, CrowdStrike, Intune, Linux and Apple pages.

Approved customers can view their tickets, plan details, projects and shared documents, create tickets with attachments and enable TOTP.

We focus on business clients in German-speaking markets. Remote work, possible on-site visits and effort are agreed per engagement.

Readable, representative samples are tested in the pilot. Faint text, handwriting and irregular tables increase review needs; OCR is not error-free.

Document processing alone does not approve payment. Posting or payment integration needs separate permissions, checks and explicit agreement.

Existing APIs, import formats and controlled exports are checked first. Replacing the ERP is not a prerequisite; a reliable handoff is.

Data scope, access, processing location and deletion are agreed beforehand. Unneeded details stay out of the workflow; production samples require approval.

Search retrieves relevant passages before the model drafts an answer. Sources remain checkable, but a citation alone does not guarantee a correct conclusion.

Yes. Checks must apply at retrieval, not only to the visible answer. Role changes and blocked sources therefore form part of acceptance testing.

Yes. Source checks, uncertainty signals and expert review reduce risk. Questions without reliable evidence require a defined refusal or escalation path.

Content owners decide approval and freshness. Synchronisation, indexing and quality checks are technically agreed and do not replace that responsibility.

Dispatchers retain critical job decisions. Rules can support suggestions and notifications; qualifications, safety and exceptions remain explicitly governed.

An agreed follow-up requests required details. The case remains visibly incomplete rather than proceeding with guessed equipment or customer data.

Suitable APIs or imports may connect the existing application. Availability, permissions and offline behaviour must be checked before commitment.

Failures are logged and assigned. Retries, duplicate checks and manual handling have explicit rules so jobs do not silently disappear.

From agreed sources and documented formulas. AI may draft commentary but must not generate measurements that cannot be traced to a source.

Where statuses, periods and exceptions share definitions. Differences in SLA rules or data collection must remain visible.

Not necessarily. Existing tools, audience, data paths and licences determine the output. A scheduled export may suit a limited need.

Checks cover required values, freshness and expected volumes. A failed retrieval is flagged and must not be released as a complete current report.

Yes, if access, licensing and permitted processing fit. A new mailbox is not inherently required.

Business owners approve rules and escalation. Uncertain or safety-critical classifications remain with a responsible person.

Only eligible messages meeting agreed criteria. Duplicates, spam and unclear content need a separate review path.

Discovery checks existing features, interfaces and licences. Any additional services, costs and processing requirements are identified before the pilot.

Not every task warrants automation. Low volumes, exceptions or poor data may favour manual handling, such as a few invoices monthly.

A named person reviews and approves relevant suggestions. An extracted order amount, for example, needs human confirmation before binding processing.

Results use changeable assumptions. A pilot measures the actual processing time, errors, queries and operating cost of the chosen workflow.

Start with a bounded pilot. Agree inputs, decision points and success criteria before considering broader rollout.