The scenario
A customer calls to check a request status, update a contact detail and ask one policy question, three intents in one call, in Arabic, with background noise.
Voice AI demo
A voice agent that holds the Saudi register, and hands over cleanly
A live call on a real service scenario: Arabic and English, interruption handling, verified identity, a completed transaction, and a transcript with the action trail attached.
In Saudi Arabia
What turns the Voice AI demo into a production system
A working demo and a system a Saudi reviewer will sign off are two different things. These are the questions asked before go-live, and what we hand over against each one.
AuthorityWhat it requiresWhat we buildWhat you keep
SDAIA, PDPL
Lawful basis for every personal-data field the demo touches, before it runs on real records.
Field-level data map produced during scoping, sanitised extract used until that map is signed off.
Data map and lawful-basis note the client's DPO can file before production access is granted.
NDMO
Data classification and quality controls on the sources the demo would depend on in production.
Source-by-source classification, with quality thresholds agreed before the pilot widens.
Classification map and a quality gate report your data office keeps.
NCA ECC
Access control, logging and read/write boundaries clear before wider rollout.
Entitlements scoped per role, every action logged, write paths gated behind approval.
Control matrix and an action log your risk function can replay.
Where it runs
The demo itself can run on a sanitised or representative extract, but where the data cannot leave a system, the build happens inside your environment inside Saudi Arabia, not ours.
Arabic, as it is spoken here
Arabic questions and documents are handled as first-class input, not translated to English before processing, including organisational and Gulf-dialect terms.
Who it is for
Contact centre, service and citizen-channel owners measured on handling time and first-contact resolution.
Sample workflow
The chain of work behind the answer
1
Call and identify
- Arabic or English detected automatically
- Identity verified before any private data
- Consent and recording notice
2
Understand
- Multiple intents in one utterance
- Interruption and barge-in handled
- Names, IDs and numbers confirmed back
3
Resolve
- System lookup for status and records
- Policy answer from governed retrieval
- Update written through a permissioned tool
4
Escalate
- Handover with full context, not a restart
- Sentiment and risk triggers
- Agent sees the transcript instantly
5
Evidence
- Transcript, summary and disposition
- Action log for anything changed
- Quality scoring on a sample
Every call ends with a written record of what was said and what was changed, which is what makes voice auditable.
What you will see
Two examples, start to finish
“أبغى أعرف وين وصل طلبي، وغيّروا رقم جوالي، وهل الرسوم مستردة؟”، three intents, one breath.
- Behind it
- Intents separated → identity verified → status read from the case system → number update written after confirmation → refund policy retrieved with its clause.
- What comes back
- All three handled in one call, a confirmation SMS triggered, and a transcript with the changed field and the clause quoted.
A caller who becomes frustrated and asks for a person.
- Behind it
- Escalation trigger fires immediately → context package assembled → routed to the right queue with priority.
- What comes back
- A human picks up already knowing the name, the case and what was attempted, no repetition.
Deliverables
What you keep after the session
Every item below is an artefact your team can reuse, whether or not the engagement continues.
A callable voice agent
One scenario, on a test number, in Arabic and English.
Call recordings and transcripts
The demo calls, with the action trail beside each one.
Language quality review
How the agent handled register, names, numbers and Hijri dates.
Containment and escalation model
Which intents to automate first and which must stay with people.
Session agenda
- You call the agent yourself, in Arabic, and try to break it
- Interruption, background noise and a mis-heard number
- Escalation to a human with full context
- Reading the transcript and action log after the call
What we need from you
- One service scenario with its current script or process
- A test endpoint or sandbox record set
- The policy content the agent may quote
Where data cannot leave a system, the demo is built inside your environment instead of ours.
FAQ
Frequently asked questions
Which Arabic does it speak?
Saudi-appropriate modern standard Arabic by default, with dialect handling on the listening side so callers can speak naturally.
Can it complete transactions or only answer?
It completes transactions through permissioned tools, with confirmation read back before anything is written.
Where do recordings live?
Inside your approved environment, under the same residency and retention rules as your existing contact centre data.
Topic cluster
Explore the topic cluster
Curated links across our Saudi hubs, sector pages and long-form blog research.