Back to Projects
AutomationClient Project

AutoFlow

Turning WhatsApp chaos into an organised, self-running order pipeline

2024AutomationPython, FastAPI, PostgreSQL, WhatsApp Business API, Celery

30+ hrs/week recovered · 10x faster order processing · Zero missed orders since launch

The Problem

Orders came in from resellers across multiple WhatsApp groups, phone calls, and DMs. The ops team spent every morning copying details from screenshots into spreadsheets, updating stock counts by hand, and calling dispatch to confirm what went out the day before. Errors were daily. Any growth would make it worse.

What We Built

AutoFlow intercepts incoming WhatsApp messages from registered vendors, parses order details, records them into a central database, and routes them to the right handler — without a human in the middle. Dispatch gets an automated morning summary. Exceptions are flagged directly to a supervisor via WhatsApp. The ops team went from managing chaos to reviewing a clean daily report.

The Outcome

Eliminated 30+ hours/week of manual order reconciliation; processing time cut from 40 minutes per order to under 4

What We Learned

In the Nigerian market, WhatsApp isn't just a communication tool — it's infrastructure. We initially tried to move the client to a web dashboard for order intake; they didn't want it. When we built around WhatsApp instead of against it, adoption was immediate and training time dropped to almost nothing.