← All work
Visit fibertime.com →
Featured Case StudyUX Design · QA LeadWhatsApp · 2025–2026

fibertime™ WhatsApp
Chatbot
& Activation Flow

800–1,000 router activations daily across South Africa — all through a WhatsApp bot with no app download. I redesigned core flows, replaced a broken location system, and shipped structured forms that cut manual entry errors.

Scroll to explore
·UX Designer & QA Lead
01Problem Framing

High-volume activations
on unreliable tooling

Field agents activate routers in high-density communities across South Africa. When location fails and data is typed freehand into chat, activations slow down and errors reach back-office systems.

Problem 01 — Location accuracy
OnMap assigned agents to the wrong address
The existing location system was unreliable in high-density areas — field agents regularly arrived at incorrect addresses, wasting time and delaying activations.
Problem 02 — Data entry errors
Customer details typed freehand into chat
Agents manually typed names, phone numbers, and contact details as WhatsApp messages — no structure, no validation, and errors that reached back-office systems.
Problem 03 — Accessibility gap
The tool wasn't built for the people using it
Field agents and customers needed a tool that worked on any Android without downloading an app. WhatsApp meets them where they are.
Chatflow designs
fibertime WhatsApp chatflow wireframes

Chatflow wireframes designed for the new WhatsApp activation bot

Final output
fibertime chatbot final result

Actual output — the live WhatsApp bot built from these designs

02Design Solution

One guided form instead
of twenty chat messages

Key insight: replace sequential free-text questions with a native WhatsApp form template — capturing all customer information through a single, guided form instead of an open-ended back-and-forth.

Try it — toggle the flow
fibertime™ Bot
online
What is the customer's full name?
john smith
Please enter their phone number.
082 555 1234
Enter alternate contact number.
0825551234
Confirm the address.
12 main rd jhb

No validation — typos reach back-office systems

Location-Based Assignment
Agent shares live location; system displays nearby homes on a map for efficient routing.
Home Assignment via FTM API
Agent selects a home from the map; system assigns it through the FTM API, syncing with back-office.
Structured Data Capture
Captures customer name, phone numbers, and contact details during activation — no paperwork.
AI-Powered Device Capture
Uses AI OCR to extract DR and Serial Numbers from a router photo, eliminating manual entry errors.
03User Flow

End-to-end activation
flow

Every step mapped — from live location share through structured data capture to device OCR and activation complete.

STARTShare Live LocationView Nearby Homes on MapSelect Home → FTM API AssignComplete Structured FormPhoto Router → OCR (DR / Serial)Confirm & Submit ActivationActivation Complete ✓
Data capture chatflow
fibertime data capture WhatsApp chatflow

A separate WhatsApp bot data capture chatflow designed for fibertime — used by field agents to capture data on houses across South Africa.

04Technical Artifact

Release discipline

Every release follows a structured QA protocol before going live.

QA process — every release
01Users informed of upcoming changes before each release
02Client demo with hands-on testing opportunity
03Deployment ticket created with full update documentation
04Developer and product documentation updated
05All flows tested on UAT number before go-live
06Figma flows updated to reflect the shipped state
07Regression testing completed across all four bot flows
05Skills Demonstrated

The Dev/PM edge
in this work

Conversational UX Design
Structured bot flows for low-literacy, high-volume field use — replacing open chat with validated form templates.
QA & Release Discipline
UAT testing, regression across four bot flows, and Figma sync on every release — treating QA as a product deliverable.
Systems Thinking
Replaced a broken location system, integrated FTM API home assignment, and added OCR for device capture — end to end.
FigmaFigJamDjangoTwilioPythonGitHubBitbucketWhatsApp Business API
Featured Case Study
See also: Yaga Proactive Project →fibertime™ WhatsApp Chatbot