Skip to content
All work
E-commerce operations · SaaS

Operately

An operations platform for brands that sell online and ship cash on delivery — it confirms every order on WhatsApp, answers Messenger and Instagram with AI, and shows why orders come back.

Operately
Year
2026
Region
Egypt
Platforms
Web · WhatsApp · Messenger · Instagram
Stack
React · .NET · MongoDB · Azure · OpenAI
Context

Where things stood.

Operately is a KBLabs product, built for online brands in Egypt that sell through Shopify and deliver through a courier, with the customer paying at the door. In that model a sale is not won at checkout. It is won when the parcel is accepted, and every order that comes back has already cost a shipment both ways.

The work around each order is spread across tools that do not talk to each other: the Shopify admin, the courier's dashboard, a WhatsApp phone passed around the team, and the brand's Facebook and Instagram inboxes. Nobody sees the whole order, and nobody can say for sure why last month's returns happened.

Challenge

What we were asked to solve.

Confirming an order before it ships is the cheapest way to cut returns, but done by phone it does not scale and is rarely done for every order. Customers also write with questions at all hours on Messenger and Instagram, and a slow answer becomes a cancelled order.

The data that explains returns already exists — the courier records a reason for every failed delivery — but it lives in another system, in the courier's own wording, and is never set against the confirmation, the product, the size or the city. And whatever automates the conversation has to know when to stop and hand the customer to a person.

Approach

How we built it.

We designed and built Operately end to end. A store connects Shopify, its Bosta account and its own WhatsApp Business number, and every new order enters a confirmation flow: a WhatsApp message with buttons to confirm, cancel or change details, a reminder and a last call for anyone who has not answered, and an "on its way" message when the courier picks the parcel up. Messages respect the store's contact hours, and the editor warns when wording would make Meta file a message as marketing, which costs more.

An AI assistant answers the store's Facebook Page and Instagram direct messages from the order's real facts and the store's own policies. The store decides when it hands over to a person; a team member who replies takes the conversation over, and presses Done to give it back.

Every order carries a risk score from 0 to 100, built from payment, the customer's record on that phone number and, once there are enough finished orders, the return rates of its product, city and weekday. Reports count finished orders only, group returns by the courier's own reasons with a suggested next step for each, time every stage of delivery, and compare the return rate of confirmed and unconfirmed orders — so a store can see whether confirming works.

Key features

What shipped.

  • Confirmation on WhatsApp, from the store's own number

    Confirm, cancel or change details in one tap, with a reminder and a last call for anyone who stays silent. Templates go to Meta in Arabic and English together, and the cost of each message is estimated before it is sent.

  • An AI assistant that knows when to step back

    It answers Messenger and Instagram from the order and the store's policies. Hand-overs wait in one inbox, most urgent first, and card numbers typed into a chat are removed before anything is kept.

  • Shopify and the courier, in one order

    Orders arrive from Shopify, deliveries from Bosta through webhooks, and each delivery is matched to its order. The team sees one timeline — what was ordered, what the customer said, every attempt at the door.

  • A risk score on every order

    From 0 to 100, with the reasons shown. Product, city and weekday only count once the store has enough finished orders for them to mean something.

  • Returns, explained

    Returns grouped by the courier's reasons, by product and by size, with what to do next for each. Exchanges are kept apart from returns, and stage timings show where delivery gets stuck.

Outcomes

What shipping looked like.

WhatsApp, Messenger, Instagram
3 channels
Before the courier picks up
Confirm first
Risk score on every order
0–100
App, guides and messages
AR + EN
Tech

What's under the hood.

Frontend
React single-page app, Arabic and English, light and dark
Backend
.NET API on Azure Container Apps, MongoDB Atlas
Messaging & AI
WhatsApp Business Platform, Messenger Platform (Facebook and Instagram), OpenAI
Integrations
Shopify app and webhooks, Bosta API and webhooks
Security
Firebase Authentication, credentials encrypted at rest, isolated stores with roles, strict CSP
Next project

Dokany

Household cleaning essentials from the factory to the door — an Arabic-first Shopify store for Egyptian families.

Have an RFP? Have an idea? Let's build it.

We reply to every inquiry within one business day. Tell us about your project, timeline, and what success looks like.