Manual QA doesn't scale
QA teams listen to 2–5% of calls. The other 95% — including the risky ones — go unheard, unscored, uncoached.
95%of calls never listened to
AI CALL CENTER QA & INTELLIGENCE
AI that listens to
AI that listens to
100% of your calls
in Arabic.
KASHEF transcribes every call, scores it against your QA scorecards, and alerts supervisors the moment it matters — so leadership sees the full picture, not a 3% sample of it.
Now onboarding design partners in Jordan, Qatar, UAE & KSA
LIVE
Call #4821 · Billing inquiry
Agent: Yousef S.
✓ Greeting +10
✓ Identity verified
✓ Empathy detected
✓ Mandatory disclosure
0
QA Score
Retail Banking Card · v3
PII redacted · card number masked
Built for the region's most regulated contact centers
Banking & FinanceTelecomInsuranceGovernmentBPO & OutsourcingCar dealerships
100%of calls transcribed & scored — not a sample
2–5%typical coverage of manual QA today
4Arabic dialect families supported at launch
24/7monitoring, alerting & reporting
The Problem
You can't run quality on a 3% sample
Most MENA contact centers still quality-check calls by hand. The math never works — and the risk never sleeps.
Compliance risk goes unseen
One missed mandatory phrase can mean a fine, a formal complaint, or a regulator's letter — discovered weeks too late.
1missed phrase is all it takes
Coaching arrives late
When reviews take days, feedback corrects last week's behavior — not the call happening right now on the floor.
Daysof lag before feedback lands
The Platform
One pipeline: from voice to verdict
Five capabilities, one Arabic-first engine. Explore what KASHEF does with every single call.
Call library100% coverage
#4821Yousef S.06:12JDبدي أستفسر عن فاتورة الشهر الحالي…94
#4820Sarah M.04:48EGعايز أفهم الرسوم اللي اتخصمت دي…88
#4819Ahmad K.09:31KSAأبغى أعرف ليش ما وصلتني رسالة التفعيل…76
#4818Leen A.03:05LEVعم اتّصل بخصوص الاشتراك، في مشكلة…97
1,284 calls transcribed today — 100% coverage
Scorecard · Retail Banking InboundAuto-fail triggered
✓Greeting & opening10 / 10
✓Identity verification15 / 15
✓Empathy & tone8 / 10
✕Mandatory disclosure0 / 20
✓Resolution & close12 / 15
Result
Auto-fail overrides weighted total — flagged for supervisor review.
Live monitor LIVE
Normal04:12
Yousef S.Billing · JDEscalated07:44
Ahmad K.Compliance phrase missedSentiment drop02:58
Leen A.Churn language detected14:02Supervisor Firas notified — Call #4819
13:57Retention script pushed to agent console
Leadership dashboardThis week
Calls8,941
Avg QA87.2
Alerts31
Coverage100%
QA trend · 7 days
Top agents this week
LA97
YS94
SM88
AK76
AI coaching · Call #4819
From the transcript
"يعني ما في حل غير إني أروح الفرع؟ طيب ليش ما انوضح لي هذا من البداية…""So there's no solution but to visit the branch? Why wasn't this clear from the start…"
AI coaching note
- Set expectations on channel limits earlier in the call
- Identity handling was strong — keep this pattern
- Offer the callback option before escalation
The Core Differentiator
Arabic isn't a feature. It's the foundation.
Global platforms bolt Arabic on as an afterthought. KASHEF was built from the first line around how your customers actually speak — in the dialect, with the idioms, at real call-center speed.
If your AI can't tell «بدي» from «أبغى» from «عايز» — it can't QA your calls.
Dialect-native models
Jordanian, Gulf, Egyptian and Levantine — trained on real contact-center speech, not news anchors.
Industry vocabulary
Custom term lists for banking, telecom and insurance — so «تسهيلات» and «باقات» score correctly.
PII redaction in Arabic
National IDs, card numbers and addresses masked in transcripts, exports and dashboards.
Sample utterance
Transcription
Model confidence
96.8%
How It Works
From first call to first insight, in four steps
01
Connect your calls
SIP, PBX or existing recordings — KASHEF plugs into your telephony stack without replacing any of it.
02
Transcribe & score every call
Arabic speech-to-text, then your scorecards applied automatically. 100% coverage, zero sampling.
03
Supervisors alerted in real time
Compliance misses and sentiment drops escalate instantly to the right person on the floor.
04
Leadership gets the full picture
Dashboards and scheduled reports in Arabic and English, drill-down ready for the board.
By Industry
Tailored to how your industry is measured
Scorecards, alerts and dashboards shaped around each vertical's compliance and CX reality.
Banking & Finance
Central-bank call-recording compliance, mandatory disclosure tracking, and early signals of mis-selling or fraud — scored on every call, every day.
Regulatory QATelecom
High-volume scoring that keeps pace with thousands of daily calls, churn-intent detection, and repeat-call driver analysis from transcripts.
Churn & CXInsurance
Claims-call accuracy, mis-selling risk detection, and policyholder experience scoring across Arabic-dialect conversations.
Claims accuracyGovernment
Arabic-first citizen-service quality with in-region data residency — built for how public-sector lines are actually audited.
Data residencyBPO & Outsourcers
Multi-client scorecards, per-program dashboards and white-label reporting — one platform across every account you run.
Multi-client
Why KASHEF
Why teams choose KASHEF
| Manual QA | Generic global platforms | KASHEF. | |
|---|---|---|---|
| Call coverage | 2–5% sampled | Varies — often sampled | 100% of every call |
| Arabic dialects | Only as good as the listener | Limited, or MSA-only | Jordanian · Gulf · Egyptian · Levantine |
| Time to insight | Days to weeks | Hours | Seconds to minutes |
| Deployment | — | Months of integration | Weeks, telephony-agnostic |
| Support | — | Global ticket queue | MENA-based team, Arabic & English |
| Data residency | Your in-house systems | Rarely in-region | In-region & on-prem options |
Security & Compliance
Enterprise trust, by design
Call recordings are among the most sensitive data an organization holds. KASHEF treats them that way — honestly, with what's live today clearly separated from the roadmap.
Available
Data residency
Recordings and transcripts hosted in-region, with on-premises deployment options for regulated and government sectors.
Available
PII redaction
Card numbers, national IDs and addresses masked across transcripts, exports and dashboards — automatically.
Live + roadmap
Access control
Role-based permissions with a full audit trail today; SSO/SAML and LDAP on the near-term roadmap.
Available
Encryption
TLS in transit and AES-256 at rest — across audio, transcripts, scores and exports.
Available
Audit logging
Every administrative action logged, searchable and exportable — built for the auditor's questions.
Roadmap
Compliance posture
Alignment with regional regulator expectations in progress; ISO 27001 certification on the roadmap — stated plainly, not promised prematurely.
We publish what is live today versus what is in progress. Enterprise buyers verify claims — we expect them to.
Pricing
Priced for how contact centers buy
Capability-based tiers. Final quotes are shaped by seats, call volume and deployment model.
Starter
Up to 20 agents
- Core quality engine
- Full Arabic transcription & QA scoring
- Standard dashboards & reports
- Email alerting
- Business-hours support
Most popular
Professional
20 – 100 agents
- Everything in Starter, plus
- Live monitor & escalation workflows
- AI coaching notes per call
- API access & CRM integrations
- Priority support
Enterprise
100+ agents & regulated industries
- Everything in Professional, plus
- In-region / on-prem data residency
- Custom dialect tuning
- Multi-client scorecards (BPO)
- Dedicated success manager
Annual contracts · pilot programs for qualifying teams. Every quote starts with a scoping call in Arabic or English.
How long does implementation take?
Connect-to-first-score typically takes weeks, not months — depending on your telephony setup. Your team hears real scored calls within the pilot's first days.
Do you require long commitments?
Production contracts are annual, but pilots start shorter so you can prove value on your own calls before committing.
Is there a minimum size?
Pricing scales with agent seats — reach out from around 20 seats upward, and we'll scope the right tier together.
Illustrative Outcomes
What KASHEF changes
These are modeled scenarios, not customer claims. Real pilot results will be published here as design-partner programs conclude — our pre-launch status, stated honestly.
Retail bank · Compliance QAFrom sample to census
3% → 100%
call coverage
QA coverage expands from a manual sample to every call; missed mandatory disclosure phrases surface the same day — not in next month's audit.
Telecom operator · Churn signalsEvery transcript scanned
100%
of calls scanned for churn language
Churn-intent phrases are clustered across all transcripts, and flagged calls are routed straight to the retention desk while context is fresh.
Government center · Citizen experienceArabic-first, audited
Real-time
Arabic QA across citizen lines
Citizen-service quality scored natively in Arabic across every line, with supervisor dashboards per service and in-region data residency.
ROI Calculator
See it on your own numbers
Move the sliders — the impact estimate updates instantly. Then bring the numbers to a demo.
Typical QA teams sample 2–5% of calls and spend 8–15 minutes per manual review. Adjust to match your floor.
Calls manually reviewed today750
Calls reviewed with KASHEF25,000
Additional calls covered monthly+24,250
QA listening hours freed / month≈ 125
At-risk calls going unheard today~364
Estimates based on your inputs, assuming ~1.5% of calls carry a compliance exposure. Request a demo for an assessment on your actual calls.
Get the full assessment
About KASHEF
Built in Amman for MENA contact centers
KASHEF exists because enterprise QA tools were never built for Arabic. We're building the category — transcription, scoring, alerting and coaching — around how the region's customers actually speak, and how its regulators actually audit.
Tareq Kasabi
Founder & CEO
Customer Experience and Contact Center Operations Leader with 18+ years of progressive experience.
View full profileClose profile
Customer Experience and Contact Center Operations Leader with 18+ years of progressive experience across telecommunications, non-profit organizations, and technology-driven environments. Proven track record of leading customer-facing operations, optimizing service delivery, developing high-performing teams, and driving measurable business outcomes through operational excellence, performance management, and customer-centric execution. Experienced in managing contact center operations, customer service functions, workforce performance, quality assurance, process optimization, and digital transformation initiatives. Recognized for delivering sustainable operational improvements, leading organizational change, and translating business objectives into practical execution through data-driven decision-making and strong leadership. Combines expertise in Customer Experience, Service Excellence, KPI Governance, Team Development, and Contact Center Leadership with growing specialization in AI-powered customer intelligence and operational automation solutions.
Eng. Mohammad Ghoqa
Co-founder & CTO
Full-stack software engineer focused on scalable web, cloud, and AI-powered software solutions.
View full profileClose profile
Eng. Mohammad Ghoqa is a Full-Stack Software Engineer with expertise in designing and developing scalable web applications, cloud-based platforms, and AI-powered software solutions. He has extensive experience across the full software development lifecycle, from system architecture and backend engineering to modern frontend development and deployment.
With a strong foundation in software engineering and a passion for emerging technologies, Mohammad specializes in integrating Artificial Intelligence into business applications, enabling organizations to automate processes, extract actionable insights from data, and deliver intelligent user experiences.
His technical expertise includes full-stack application development, API design and integration, database architecture, cloud deployment, DevOps practices, and AI-driven solution development. He is committed to building secure, high-performance, and maintainable systems that align technology with business objectives.
As a technology leader, Mohammad combines engineering excellence with a problem-solving mindset, continuously exploring innovative approaches to deliver impactful digital products and scalable software platforms.
Request a Demo
See KASHEF on your own calls
Tell us about your operation — we'll arrive prepared with a demonstration scored on calls like yours, in Arabic or English.
We reply within one business day
A real person from the MENA team — not an autoresponder.
Discovery call in Arabic or English
Your QA process, telephony stack, and compliance constraints.
Demo scored on sample calls from your industry
Transcription, scorecards, alerts and dashboards — live.