Automation / AI / Business system
Enquiries should not disappear inside your inbox.
An automated lead-handling workflow that understands enquiries, collects missing details, updates the lead system, alerts the owner and follows up automatically.
Portfolio project by AdamSofi · Demo mode
Enquiry workflow
RunningCustomer enquiry
Hi, saya nak buat website untuk company renovation…
AI understands request
Collect information
Update lead system
Notify owner
Automatic follow-up
Owner action needed
—
Manual vs automated
Seven manual steps become one workflow.
The enquiry itself was never the hard part. The repetitive work that follows it is what costs the owner a day.
Before
7 manual steps- Read enquiry
- Ask questions
- Wait for reply
- Copy details
- Update spreadsheet / CRM
- Notify owner / team
- Remember to follow up
- Slow first response
- The same questions, every time
- Incomplete information
- Messages buried in chats
- Forgotten follow-ups
After
1 automated flow- Enquiry
- Qualification
- Structured lead
- Owner alert
- Automatic follow-up
- Reply starts immediately
- Only the missing details are asked for
- Every lead arrives structured
- Important leads surface first
- Follow-ups are scheduled, not remembered
The workflow
Six stages, none of which need the owner.
Select a stage to see what goes in, what comes out, and which manual step it removes.
Stage 02
AI understands request
Keyword and pattern rules read the message and detect what the customer is actually asking for, along with the type of business behind it.
- Input
- Enquiry text
- Output
- Intent · business type · features · budget · timeline
- Manual step removed
- Manually interpreting every message
Core capabilities
What the system takes off the owner's desk.
Each capability maps to a step that would otherwise be done by hand after every single enquiry.
Enquiry understanding
Detects the service, business type, requested features, budget and timeline from a message written in plain Malay or English.
Structured customer information
Every enquiry becomes the same shaped record, so leads can be compared, filtered and searched instead of scrolled through.
Fewer repetitive questions
The system asks only for what is actually missing. The customer answers three or four short questions, not a form.
Clear lead priority
A transparent points rule assigns Hot, Warm or Cold — and the lead page always shows why that score was given.
Owner notifications
High-priority leads reach the owner immediately, with enough context to act without opening anything else.
Follow-ups that are scheduled
A follow-up is created the moment a lead qualifies, and stays on the dashboard until it is done or cancelled.
Owner visibility
One dashboard shows what needs attention now, what is in the pipeline, and what the automation has been doing.
Owner dashboard
The owner opens one screen and knows what to do.
Not vanity analytics — a short list of decisions, ranked. Leads created in the live demo appear here immediately.
Overview
Demo data12
last 7 days
4
open
3
today or overdue
28
by automation
Needs attention
Amir Renovation
HotWebsite · RM3k–RM5k
Follow-up due today
Klinik Pergigian Seri
HotAI Agent · RM5,000+
Follow-up overdue
Warisan Katering
WarmAutomation · RM1k–RM3k
Contact number not provided
Example enquiry journey
One message in. A qualified lead out.
This is the default scenario in the live demo — a renovation contractor asking about a website, in the way people actually write.
01 · The message
“Hi, saya nak buat website untuk company renovation. Nak tahu anggaran harga.”
No name, no budget, no timeline, no contact number. Typical, and unusable as it stands.
02 · What the system extracted
- Intent
- Website
- Business type
- Renovation
- Features
- Company profile, services, gallery, WhatsApp
- Budget
- RM3,000 – RM5,000
- Timeline
- Within 2 Weeks
- Contact
- 012-345 6789
03 · What happened next
Hot- Lead score calculated: 92
- Lead marked HOT
- Lead saved · #L1041
- Owner notification created
- Follow-up scheduled · WhatsApp
Owner steps requirednone
Case study
Problem, approach, system, result.
The full write-up covers how the workflow was designed and where real providers plug in.
Problem
The enquiry was not the problem. Everything after it was — reading, asking, copying, recording, notifying, remembering.
Approach
Turn an unstructured message into a structured lead first, then let rules handle the steps that always follow.
System
An enquiry interface, a deterministic processor, a qualification engine, a lead store, notifications and a follow-up scheduler.
Result
Enquiries arrive structured, important leads become visible, and follow-ups stop depending on memory.
A customer enquiry enters once, and the system handles the repetitive work that comes after it.
Run the automation yourself. It takes about seven seconds, and the lead it creates lands in the dashboard.