AI for Insurance Agency Malaysia: An Agentic AI Case Study

AI for insurance agency Malaysia, in practice: how Aimplify used agentic AI to automate claims emails, quotations, commissions and WhatsApp updates.

Running a general insurance agency in Malaysia is a juggling act. Your agents are out meeting customers, your admin team is buried in insurer emails, and somewhere in that inbox is a claim update a customer is waiting for. This case study walks through how Aimplify helped one insurance agency client bring agentic AI into its daily operations, without ripping out the tools the team already knew.

The client is a Malaysian general insurance agency handling many motor policies alongside other general insurance lines. Like many agencies, it ran on Google Workspace, email and WhatsApp. The problem was never the tools. It was the manual work sitting between them.

The Pain Points: Where the Admin Hours Were Going

Before we built anything, we sat with the team to find the real bottlenecks. Six stood out.

Every email read and filed by hand

The shared admin inbox received a steady stream of insurer and customer emails. Staff opened each one, decided whether it was about a policy or a claim, created a Google Drive folder for the customer, saved the attachments and worked out the next step. Important, yes. But slow and repetitive.

The founder had no live view from his phone

The founder spends most of his day out selling. When a customer called asking "how's my claim?", he had no quick way to check status from his phone. He had to call the office and wait for someone to dig through emails.

Claim emails scattered, insurers chased manually

Emails about a single claim arrived in separate threads, often without a policy number. When an insurer went quiet, someone had to remember to chase them, write the follow-up and track whether they replied.

Slow lookups for motor policies

Motor policies are identified by car plate. When a customer asked "am I covered for windscreen?" or "do I have towing?", staff had to find the right policy document and read through it.

Manual quotation comparison

For every prospect, the team collected quotations from several insurers in different formats, then compared them line by line to prepare a recommendation.

Commission vouchers built by hand every month

At month end, admin compiled agent commission figures and produced payment vouchers manually: matching customers to agents, handling withholding tax, and catching duplicates and reversals.

The Workflows: Agentic AI Built on Tools They Already Use

Aimplify built a set of custom agentic workflows on top of the agency's existing Google Workspace and WhatsApp, using serverless cloud functions and an AI model to do the reading, sorting and drafting. Here is what each one does.

Workflow A: Commission voucher automation

This is the closest to our Aimplify Accounting agent. The admin pastes the monthly commission data into a Google Sheet. The workflow then:

• Keeps only the commission rows and ignores everything else
• Matches each customer to the right agent
• Calculates commission after withholding tax
• Skips duplicates and reversals automatically
• Generates one PDF voucher per agent in a dated Google Drive folder
• Produces a subtotal summary for the admin to review

If month-end paperwork is a pain point in your business too, our guide on month-end receipt automation for Malaysian SMEs covers the same principle for expenses.

Workflow B: Insurance email automation with a daily WhatsApp report

Every day, the workflow reads new emails in the shared admin inbox and logs one row per email thread in a sheet, saving attachments to Google Drive. The AI then:

• Classifies whether the thread is about a claim or a policy
• Pulls out key details such as the policy number, insured amount, claim amount and insurance type
• Writes a short summary and sets a status: documents needed, pending with insurer, approved or rejected
• Flags missing information so admin knows what to request

When a later reply comes in on the same thread, the same row is updated rather than creating a duplicate. Scattered emails finally sit together.

The founder receives a daily WhatsApp report with the summaries and Drive links. Now when a customer asks about a claim, the answer is already in his pocket.

Workflow C: Insurance claims automation for follow-ups, with human approval

Chasing insurers is where many agencies lose time, but it is also where tone matters. So we kept people in charge:

• Admin flags which claims need chasing
• The AI drafts a follow-up and places it in a Pending tab
• Admin reviews and approves each draft before anything is sent
• A Sent tab tracks replies and how many follow-ups each claim has had
• A do-not-contact box always blocks sending, no exceptions

The timing is a setting, not a guess. For example, the agency can choose to flag a claim if an insurer hasn't replied in 3 days.

Workflow D: AI insurance quotation comparison

Admin drops the insurer quotations into a Google Drive folder. They can be PDF, Word, Excel or even images. One click on "Run comparison" and the AI builds a side-by-side table, then produces a branded PDF ready to send to the prospect.

Because insurers change their layouts, staff can adjust the extraction hints themselves, with no developer needed.

Workflow E: An internal WhatsApp bot for insurance agents

This is the closest to our Aimplify CS agent, used internally by staff rather than customers. Policies are filed in a Drive folder per vehicle plate. Staff simply ask on WhatsApp, for example:

• "Does this car have windscreen cover?"
• "Is towing included?"
• "Any special perils or no-claim discount on this plate?"

The bot answers and can send back the policy PDF on the spot. Staff can also send new policy files to the bot, which files and logs them automatically. It runs on the agency's own WhatsApp number.

Workflow F: WhatsApp policy-expiry reminders

Before a vehicle policy expires, a WhatsApp reminder goes out. The timing is configurable; the agency set it to aim for around two months before expiry, giving customers time to plan ahead.

The Outcome So Far

The commission vouchers and the email screening with the daily WhatsApp report have been live since late May 2026. The quotation comparison PDFs, the internal vehicle policy bot and the expiry reminders were finished and tested by late September 2026, and the project is now in customer testing.

The feedback has been encouraging, and we'll keep it honest:

• The founder gave very positive feedback on the daily WhatsApp summary. Claim status is now something he sees each day rather than something he has to chase.
• The admin team confirmed the quotation comparison tool works after a short walkthrough, and they can tune it themselves.

The goal was never to cut headcount. It was to give the admin team back their focus for the work that needs a human: talking to customers, handling tricky claims and supporting the sales effort.

Human-in-the-Loop and Data Privacy

Insurance data is sensitive, so we designed for control from day one.

• Humans approve outbound messages. Claim follow-ups are drafted by AI but only sent after admin approval.
• Do-not-contact is absolute. If the box is ticked, the workflow will not send.
• Data stays in the agency's own workspace. Emails, sheets and documents remain in the client's Google Workspace and Google Drive, not in a new system they don't control.
• Transparent logs. Every email thread, follow-up and reply is visible in a sheet that staff can check and correct.
• Staff can adjust the AI. Extraction hints and settings, like reminder timing, are editable by the team.
• Anonymised case studies. We never publish client names, staff names or policy details, which is why this post doesn't either.

Is This Right for Your Agency?

Agentic AI is likely a good fit if you can tick a few of these:

• [ ] Your admin team reads and sorts insurer emails by hand every day
• [ ] You or your agents often can't see claim status while on the road
• [ ] Insurer follow-ups depend on someone remembering to chase
• [ ] Finding cover details for a specific vehicle takes too long
• [ ] Quotation comparison for prospects is done manually in spreadsheets
• [ ] Commission vouchers are built by hand each month
• [ ] You already use Google Workspace, email and WhatsApp, and want to keep them

If three or more sound familiar, it's worth a conversation. The same approach applies beyond insurance; see how we tackle rekeying in SQL Accounting invoice automation.

FAQ

Do we need to change our email or Google Workspace setup?

No. Aimplify builds on the tools you already use. In this case study, everything ran on the agency's existing Google Workspace, Google Drive and WhatsApp.

Will the AI send messages to insurers or customers without our approval?

Not for claim follow-ups. Every draft sits in a Pending tab until admin approves it, and a do-not-contact box always blocks sending. Policy-expiry reminders follow timing rules the agency sets.

Can the AI read quotations in different formats?

Yes. The quotation comparison workflow handles PDF, Word, Excel and image files, and staff can adjust the extraction hints when an insurer changes its layout.

Is this only for motor insurance?

No. The email screening, claim follow-ups, quotation comparison and commission vouchers work across general insurance lines. The plate-based WhatsApp bot and expiry reminders are tailored for motor policies.

How does Aimplify keep client data private?

Data stays in your own workspace, humans approve outbound messages, and every action is logged where your team can see it. We also anonymise all case studies.

Ready to Bring Agentic AI to Your Insurance Agency?

Every agency has its own bottlenecks. Aimplify will sit with your team, find where the hours are going and build an AI workforce around the tools you already use, with humans in control.

Talk to an AI Architect at aimplify.com.my and let's map out what agentic AI could do for your agency.