
Al Abdulghani Motors' agentic AI for sales and service. Live first on WhatsApp, in Arabic and English, connected to Salesforce and SAP S/4HANA. Built to extend to every channel.
Agentic AI
Three generations of automation. Sayyar is the third.
Menus and keywords. Breaks the moment the customer goes off script.
Understands and writes well. Cannot book, approve or change anything.
Understands the goal, acts in your systems, finishes the task, and hands over to a person when it should.
The brand
One name customers recognise on every channel. One agent behind it, powered by Sprinklr.
Arabic for the one who moves, rooted in sayyara, car. Always on the road for the customer: a booking, an approval, a test drive.
"Hi, I'm Sayyar from Al Abdulghani Motors." Same voice and name whether they message, call or comment.
A customer-facing asset Al Abdulghani Motors owns and grows: new skills and channels ship under one brand.
One Sprinklr AI+ agent, one set of skills and integrations. Every new channel reuses what phase one builds.
Why Sprinklr
AAM’s CRM team already works in Sprinklr. Putting Sayyar there too changes seven things.
Sayyar escalates; the agent picks up with transcript, intent, sentiment and what the bot tried. The customer repeats nothing.
Every escalation crosses two platforms. Partial context means the customer explains again; each new channel rebuilds the handoff.
The knowledge agents and Copilot use today becomes Sayyar’s. One update, say a new fee, reaches bot, Copilot and agents at once. Next: agent learnings feed the bot.
Bot knowledge in one system, agents’ in another. Every update made twice; miss one and bot and agent disagree.
A complaint in an Instagram comment: Sayyar replies publicly, moves to DM, resolves or hands off. Native.
Instagram, X and TikTok comments need a partner tool, most cover DMs only. Comments stay manual and off the analytics.
A customer calls in Gulf Arabic. The AI agent understands, completes simple requests and transfers to a human on the same platform.
AI voice documented as English only. Arabic calls wait for a roadmap or a third vendor.
Why Sprinklr
Analytics, feedback and campaign replies stay whole when the bot is not a separate system.
One view for management: top complaint topics across bot, calls, WhatsApp and social; where Sayyar fails; why customers escalate; sentiment trends in Arabic and English.
Bot conversations analysed in one system, everything else in another. Someone exports and merges; in practice the bot is left out of the picture.
After a bot or agent conversation a survey goes out. If CSAT drops, click through to the exact conversations behind the score, whoever handled them.
Bot surveys sit apart from agent and channel surveys. Scores in two places; tying a low score to the journey is manual work.
A renewal reminder goes to 10,000 customers. One replies “How much is the fee?” Sayyar knows which campaign it is, answers, hands off or creates a lead in Salesforce.
Campaign in one system, bot in another, agents in a third. Replies land without context, get missed or need manual routing. Each new channel repeats the chain.
Target scope
Sales first, then service operations, then upsell. Built to extend to other business units.
Qualify interest in a Toyota or Lexus model, book a test drive at the chosen showroom, create the lead and opportunity in Salesforce.
Look up the booking, offer open slots, and update the service order in S/4HANA.
Extra work found during service (brake pads, filters) is quoted on WhatsApp. Approval flows back to the S/4HANA job card with no human intervention.
Capture location and issue, identify the vehicle, dispatch a request, and keep the customer updated.
Offer packaged deals at the right moment, such as 40K to 120K km coverage at a bundled discount.
Journey Facilitator already sends surveys and campaigns. When the customer replies, the agent picks up the conversation instead of a queue.
Strategy
Build the agent once. Each phase adds channels, not a new bot.
Sales leads and test drives. Job approvals, reschedules, roadside. Replies to surveys and campaigns. Arabic and English.
Same Sayyar on Instagram, Facebook, X DMs and comments, plus website and app chat. Public complaints handled before they spread.
Sayyar answers the service hotline in Arabic and English. Books, approves and escalates by phone. A call can continue on WhatsApp.
How it is built
Start with the job to be done, ground the Agent, then provide the actions it can execute.
Give the Agent a name, purpose and behaviour. For Al Abdulghani Motors: Sayyar, service and sales assistant, Arabic and English.
Link the Sprinklr Knowledge Base, import documents in any format, fill gaps with custom FAQs.
Add specialised skills, macros, APIs and prompt chains. For Al Abdulghani Motors: Salesforce and S/4HANA VSS actions.
Platform and integrations
Sprinklr runs the channel and the agent. Systems of record stay where they are.
One verified AAM number. Text, buttons, lists, images, documents. More channels switch on later.
Intent and language, reasoning and tool calling, authentication by mobile number, handoff to the CRM team in the same console.
Leads, opportunities, test drives as native objects. Existing AgentForce logic reused.
Service orders, job cards and appointments read and written via APIs.


Sayyar: every Toyota and Lexus customer served in their language, on WhatsApp today and every channel next.