01
Knowledge buried across documents
People search folders and ask colleagues to locate a policy or instruction.
AI / RAG knowledge systems
Document search and assisted answers designed around your sources, access rules and questions in the agreed languages, with evidence close to the response.
Scope, timeline and pricing agreed after reviewing your needs.
A question, supported by sources
- 01Question
- 02Sources
- 03Evidence
- 04Answer
Illustrative concept, not a live system or measured result.
RAG systems
Document ingestionRetrievalSource referencesAccess controls
Who this is for
For organisations with documents scattered across folders and systems, answers that cannot show their evidence and access rules that differ from person to person.
02
An answer with no evidence
A convincing response is difficult to trust when its source is unclear.
03
Different people, different access
Useful answers must respect which documents each person may see.
What you get
Extraction and readiness review, retrieval with visible sources, an access model, evaluation in the agreed languages and an agreed owner for refreshing the knowledge.
Source and question review
Inventory documents and representative questions before selecting a retrieval approach.
Document preparation
Plan extraction, structure, updates and ownership for the agreed source types.
Retrieval with visible evidence
Connect relevant passages to an answer experience with usable source references.
A question, supported by sources
QuestionUnderstand what the person needs to find.
Permitted sourcesRetrieve relevant passages within the access rules.
Evidence reviewSelect useful context and notice missing support.
Document ingestionRetrievalSource referencesAccess controls
Permission-aware access
Apply the agreed access model to indexing, retrieval and source links.
Evaluation in the agreed languages
Test questions in the selected language or languages, document quality and unsupported answers.
A maintained knowledge workflow
Define source refresh, evaluation checks and who reviews reported issues.
How we work
Review the sources and the questions people actually ask before designing ingestion and retrieval, then test both supported answers and the cases where the system should decline.
Inventory knowledge
Review the document library, owners, permissions and questions people actually ask.
Know the source material.
- Document sample
- Source owners
- Question set
Prepare retrieval
Build extraction and indexing around the source quality, language and access model.
Useful context starts here.
- Extraction checks
- Retrieval design
- Permission mapping
Evaluate answers
Review retrieved evidence, answer support and cases where the system should not answer.
Check the answer and its basis.
- Retrieval relevance
- Citation checks
- No-answer cases
Keep it current
Pilot the experience and agree how changed documents, access and feedback are handled.
Knowledge needs an owner.
- Source refresh
- Feedback review
- Release evaluations
Explore the possibilities
Illustrative concepts from our demo portfolio. They show possible experiences, not commissioned client projects, platform-specific implementations or measured results.
Sources within reach
The knowledge-assistant concept pairs a question field with document cards, illustrating how references can stay visible in the experience.
Find the right learning material
The learning-platform concept provides a setting for exploring assisted discovery across courses and approved learning resources.
PLANS & INCLUDED BENEFITS
The assessment reviews readiness and ends with an architecture and estimate, not a deployment. The pilot covers one language and one source; the department plan is bilingual with scheduled refresh. All plans require machine-readable documents, exclude OCR and apply the document or page limit, whichever is lower.
Defined scope
Knowledge assessment
Check whether your content is ready for a retrieval system.
Starting from
6,000SAR
/ project · excl. VATOne-time project fee. Support beyond the included period is quoted separately.
Included in your plan
- Review up to 100 sample documents
- Map one knowledge use case and access rules
- Create a 30-question evaluation set
- Retrieval architecture and build estimate; no deployed system
Knowledge-readiness workshop included
Included in the starting price. Scope and limits are listed below.
Recommended
RAG pilot
Test grounded answers against one controlled knowledge collection.
Starting from
30,000SAR
/ project · excl. VATOne-time project fee. Support beyond the included period is quoted separately.
Included in your plan
- Up to 500 text documents or 5,000 pages, whichever is lower
- One ingestion source and one language
- Private test interface with source citations
- 50-question retrieval and answer evaluation
Team handover session included
Included in the starting price. Scope and limits are listed below.
Defined scope
Department knowledge
A bounded bilingual knowledge system for one department.
Starting from
70,000SAR
/ project · excl. VATOne-time project fee. Support beyond the included period is quoted separately.
Included in your plan
- Up to 2,000 text documents or 20,000 pages, whichever is lower
- Up to 2 sources with scheduled refresh
- Arabic/English, one sign-in provider and up to 3 access groups
- 100-question evaluation and operational handover
30-day defect fixes included
Included in the starting price. Scope and limits are listed below.
Starting prices in SAR, excluding VAT where applicable. Final scope and price are confirmed in your proposal. Model/API usage, hosting, software subscriptions and major data cleanup are separate. Scope assumes client-provided access to documented APIs and approved data. No foundation-model training or autonomous high-risk decisions. Handover is one 60-minute session; included defect fixes cover the delivered scope for 30 days. Scanned-document OCR and guaranteed answer accuracy are not included; source documents must be machine-readable. The readiness workshop lasts 60 minutes.
Before we start