Solution · Customer Service & Engagement

One approved answer, wherever the question is asked

Knowledge fails for a structural reason: it is authored as documents but consumed as questions. This solution places a single governed knowledge base behind customers, agents and AI, and answers the question rather than returning the document.

Explore this solutionSee X.Orbit
Knowledge & Self-ServiceTalk to X-Shift

Overview

Governed once, served everywhere

The value lies in the governance as much as in the search.

X-Shift establishes one knowledge base with a defined owner, an approval workflow and version control, and serves it to every audience that requires it: the public self-service portal, the agent workspace and the AI agent. Because all three draw on the same approved content, the answer is consistent regardless of where the question is asked. Where a request follows a defined procedure, the platform guides the user through it step by step rather than supplying the document that describes it.

The knowledge store is part of X.Orbit, and the AI answering from it is X.24; both are built by X-Shift, so the approval state is a single shared record. When an article is retired once, it is removed from the public portal and from the content the AI may use. No superseded policy remains active in an index that has no owner.

Key features

Key features

01

Knowledge portal

A central portal with role-based access, serving customer, agent and employee audiences from a single content set.

and 5 more below

Knowledge & Self-ServiceIllustrative
FEATURE Knowledge portal Search Knowledge portal Search Approved AI semantic search Published Ask AI & instant answers Published Results illustrative

Guided assistance

Defined procedures presented as guided steps rather than a document requiring interpretation.

Self-service

Customer-facing resolution for questions that do not require a person.

Approval & publishing

Authoring, review and approval workflow, with version management and scheduled review dates.

Governance

Defined content owners, access control, change auditing and retirement of superseded content.

Feedback & analytics

Insight into what is searched, what is found, what fails and which content requires rewriting.

In practice

A fee schedule changes on a Sunday

A fee schedule is revised and takes effect the following morning. Previously, the update reached the intranet the same week and the public FAQ a month later, while agents were informed through a team briefing that half the shift missed.

01

The article is edited once, reviewed by its defined owner and published with an effective date.

02

The portal, the agent workspace and the AI agent all serve the new version from that point onward.

03

The superseded version is retired and is no longer available as grounding for an AI answer.

This is the work performed by X.Orbit's knowledge governance. It does not write the article, and it cannot resolve a policy that two departments still describe differently; it can only make the disagreement visible.

The challenge

Three audiences, three versions of the truth

Most organisations do not have a single knowledge problem; they have three.

The customer is given a public FAQ that was last reviewed two years ago. The agent is given a shared drive and a colleague to ask. The employee is given a policy PDF and no means of determining whether it is current. All three answer the same question from different sources, so the organisation provides three different answers and cannot explain why.

Content dispersed across portals, drives, intranets and inboxes

No reliable means of identifying which version is approved and current

Agents interpreting policy documents in real time, during the call

Public self-service that addresses only the questions nobody asks

New staff reliant on colleagues rather than on content

AI grounded on content that was never approved for that purpose

Audiences

Who it serves, and with what

Who

Three audiences

01 Customers - public and authenticated self-service
02 Agents - answers surfaced inside the case
03 Employees - internal policy and procedure
04 AI agents - grounded, approved source content

Each audience is served from the same approved content estate; the difference lies in what is surfaced and in what context.

What

Content types

01 FAQs and articles
02 Policies and procedures
03 Process flows and closed process flows
04 Step-by-step guides
05 Pictorial and visual guides
06 Documents and service knowledge

Content types are governed, versioned and approved, so the source that serves one audience is the same source that serves the others.

Outcomes

What changes for the operation

Consistency is the first benefit and the one customers notice; speed follows.

01

One answer, every channel

The same question receives the same answer wherever it is asked

02

No real-time policy interpretation

Agents no longer interpret documents during the contact

03

Effective self-service

Resolves questions rather than redirecting them

04

Faster onboarding

New staff attain competence without weeks of shadowing

05

Trusted grounding

AI has a reliable source on which to ground its answers

06

Visible content gaps

Gaps are identified from search-failure data

Use cases

Questions answered three different ways

Each of these begins as a content problem and proves to be a governance problem.

01

Service eligibility explained inconsistently with X.Orbit

Situation

Eligibility rules for a public service are published on the portal, summarised differently in the call centre script and explained in a third way at the counter.

What happens
  1. Identify the customer
  2. Create and classify the case
  3. Apply SLA and owner
  4. Resolve and close
Outcome

One approved article serves the portal, the agent and the counter, so a beneficiary receives the same rule everywhere.

Commonly applied in Government and across other sectors

02

Agents interpreting policy during the call with X.Orbit

Situation

An agent reads a product policy aloud during a live call, interprets a clause that two departments read differently and commits the bank to a single interpretation.

What happens
  1. Identify the customer
  2. Create and classify the case
  3. Apply SLA and owner
  4. Resolve and close
Outcome

The clause is published in X.Orbit as an approved answer, so the agent cites an established position rather than improvising one.

Commonly applied in Banking and across other sectors

03

Preparation instructions before a procedure with X.Orbit

Situation

Patients arrive unprepared for procedures because the instructions are held in a leaflet, a call centre note and a printed clinic sheet that contradict one another.

What happens
  1. Identify the customer
  2. Create and classify the case
  3. Apply SLA and owner
  4. Resolve and close
Outcome

Preparation is delivered as guided steps from a single owned article, so the clinic and the call centre provide the same information.

Commonly applied in Healthcare and across other sectors

More ways to apply it

Admission policy questions at scale with X.Orbit

The forty questions are published as answers rather than as a policy document, and the same text serves both applicants and staff.

Education

Tariff launched to an unbriefed floor with X.Orbit

The tariff content is approved and surfaced within the agent workspace, so attendance at the briefing no longer determines the answer.

Telecommunications

Procedures held by long-serving staff with X.Orbit

The procedure is written as guided steps with a defined owner, so competence no longer depends on who remains in post.

Utilities
Explore by industry

These are illustrative applications. The content to be prioritised is determined by what customers and agents search for, not by the size of the document estate.

Common questions

What buyers ask before committing

Next

Where to go next

Test it against your worst article

Select a policy that your agents routinely misread, and observe it restructured, published to three audiences and answered as a question rather than returned as a document.

Applicable where the content causing the recurring contacts is already known.

Request a content session

Define the procedures first

Customer Service Process Design establishes the procedure itself before it is documented, which is significant where two departments describe the same service differently and both believe they are correct.

Applicable where the content is inconsistent because the process is.

Customer Service Process Design

A knowledge platform will publish an unresolved policy as efficiently as a settled one; where this is the underlying problem, we will state so.

Where this sits

Related

Runs onX.OrbitIntegrated AI knowledge management.
Applied inEducationApplicant, student and policy knowledge at volume.
Deployed asAI-Powered Customer ServiceAI that answers from this knowledge base.
Deployed asCustomer Service & Case ManagementKnowledge surfaced inside the case.

Also relevantCustomer Service Process DesignTraining & Enablement

Technology that enables it

What this is built on

These are platforms built and supported by X-Shift. Where your estate already covers one of them, the engagement becomes integration and operating model work rather than a platform deployment, and we will state this clearly.

Unified Customer Service & Engagement Platform

Interactions, cases, knowledge and service intelligence on a single platform.

Explore
Case workspace with Copilot Illustrative
QUEUE CASE-48190 Billing query CASE-48191 Complaint · SLA 3h CASE-48192 Service request CASE-48193 Renewal CUSTOMER NA Noura A. Retail banking · 4 prior cases · last contact 6 days CASE · BILLING ADJUSTMENT Disputed charge on a closed account Complaint High severity SLA 3h 42m INTERACTION HISTORY WhatsApp IN Voice call OUT Email IN AI Copilot Case summarised Four interactions, two channels Classification suggested Billing → disputed charge Knowledge retrieved Adjustment approval procedure Customer context Account closed 11 days ago Next best action Raise credit, notify customer Response drafted Ready for review Handover summary

X.Orbit: Unified Customer Service & Engagement Platform

Is this the capability you require?

Describe what is not working today, and we will advise whether this is the right answer, a smaller piece of work, or an entirely different problem.

Start a conversation