Order-issue tickets that took days now close in about five minutes
DigiHaat is a multi-category shopping app — food, groceries, beauty, electronics — built on India's Open Network for Digital Commerce (ONDC). Nanostuffs built a Salesforce Service Cloud desk for DigiHaat's Customer Service Desk (CSD), with custom objects and a live order-tracker integration that pulls order status out of DigiHaat's own systems, plus an AI-based chat agent on the DigiHaat app and WhatsApp — cutting order-issue resolution from days to about five minutes.
On an ONDC order, the answer to “where is my order” doesn't live in one place — it's split across DigiHaat, the seller's own system, and the delivery network, and every ticket used to mean chasing all three by hand.
How the CSD desk works
Video coming soon
At a glance
Client
DigiHaat — a multi-category shopping app (food, groceries, beauty, electronics) built on the ONDC network, connecting buyers to thousands of independent sellers across India
Challenge
CSD agents had no single view of an order — resolving a status query meant manually cross-checking DigiHaat's system against the seller's own fulfillment data
Solution
Salesforce Service Cloud with custom objects for ONDC orders, a live order-tracker integration, and an AI chat agent on the app and WhatsApp
Result
Order-issue resolution time cut from days to roughly 5 minutes per ticket
One order, three systems, and a CSD team stuck reconciling all of them by hand
DigiHaat is the buyer-facing app for India's Open Network for Digital Commerce (ONDC) — it lets people order food, groceries, beauty products and electronics from independent local and national sellers through a single app, rather than through one retailer's own catalog and warehouse.
That network model is DigiHaat's strength for buyers, but it's exactly what made customer service hard. Because every order is fulfilled by a different seller on the ONDC network, DigiHaat's own Customer Service Desk (CSD) couldn't answer a basic “where is my order” ticket from one screen — an agent had to step outside their support tool to reconcile DigiHaat's order record against the seller's fulfillment status, channel by channel. That manual reconciliation was slow enough that a single order-issue ticket could take days to close.
A single cart could combine grocery items from one seller with electronics from another, so a single customer complaint could actually span multiple sellers and fulfillment providers arriving through different support channels — app chat, WhatsApp, and direct contact. Agents needed one place to see the whole picture, not several tabs to piece it together.
Engagement Snapshot
- Industry
- E-commerce / ONDC Network
- Geography
- India
- Categories
- Food, Grocery, Beauty, Electronics
- Core platform
- Salesforce Service Cloud
- Integration
- Custom order-tracker + AI chat (app & WhatsApp)
Why Salesforce Service Cloud was the right core for multi-seller marketplace support
A CSD team fielding order and delivery complaints needs case management, SLA tracking, and routing across channels before it needs anything custom. Salesforce Service Cloud gives DigiHaat that foundation out of the box — a place to log a Case, route it to the right agent, and track it to resolution — regardless of which channel the customer came in on.
The build-vs-configure line was clear: standard Salesforce objects (Cases, Omni-Channel routing, Knowledge for agent macros) handled the support workflow itself. What Salesforce doesn't know natively is DigiHaat's own order state or an ONDC seller's live fulfillment status — that context lives outside the platform, and had to be engineered in as a custom integration and data model.
What Salesforce Service Cloud handles natively
- Case creation, SLA tracking and escalation for every support ticket
- Omni-Channel routing so app-chat and WhatsApp conversations reach the right CSD agent
- A shared Knowledge base and macros so agents answer common queries consistently
Where custom engineering was needed
Salesforce has no built-in concept of “Order #48213 is late because the seller hasn't dispatched it yet” — that's DigiHaat- and ONDC-specific order and fulfillment data. Getting it in front of an agent, live, on the Case they're working, meant building custom objects and an integration layer rather than relying on anything off-the-shelf.
What we built
01 · Layer
Case Management & Custom Objects
Standard Salesforce Cases extended with custom objects that model ONDC orders, sellers and fulfillment status, so an agent's view of a ticket already includes the order context, not just the complaint text.
Salesforce Service Cloud02 · Layer
Order Tracker Integration
A custom integration reaches into DigiHaat's own order system to pull live order and fulfillment status straight into the agent console, replacing the manual cross-checking agents used to do outside Salesforce.
Custom REST API integration03 · Layer
AI Chat Agent — App & WhatsApp
An AI-based chat agent handles first-line queries on live chat and WhatsApp, resolving routine order questions on its own and handing off to a CSD agent in Salesforce when a case needs a human.
AI Chatbot + WhatsApp Business APIGiving agents one accurate view of an order that doesn't actually live in Salesforce
DigiHaat is a buyer app sitting in front of an open network of independent sellers — order and fulfillment status is generated and owned outside Salesforce, in DigiHaat's own order system and, ultimately, in each seller's own operations. None of that is data Salesforce starts out knowing.
The obvious shortcut — having agents open a separate order-tracking screen alongside Salesforce, or refreshing a periodic data export into a custom object — either kept agents jumping between tools or left the order status stale by the time a customer asked about it, which defeats the point of a faster resolution.
The fix wasn't to move order data into Salesforce — it was to build a live order-tracker integration that reaches out to DigiHaat's order system in real time and surfaces exactly what an agent needs, right inside the Case they're already working.
When a Case is opened for an order, the integration layer calls DigiHaat's order system, resolves the current status against the custom order object, and renders it in the agent's console alongside the ticket — so the agent sees the same “where is my order” answer a customer would see, without leaving Salesforce or asking the customer to wait while they check somewhere else.
Before
- Agents manually cross-checked order status outside Salesforce for every ticket
- Resolving a single order-issue ticket could take days
- App-chat and WhatsApp queries were handled as separate, disconnected threads
After
- Live order status appears directly on the Case inside Salesforce
- Order-issue tickets now resolve in about 5 minutes
- AI chat agent handles routine queries on app and WhatsApp before they ever reach an agent
How we delivered it
We scoped the build around the CSD team's actual ticket flow first, then layered in integration and AI — de-risking each phase before building on top of it.
Discovery
Mapped how CSD agents actually handled an order-status ticket end to end — every system they opened and every channel a query could arrive on — to scope what Salesforce needed to absorb.
Service Cloud Foundation
Configured Salesforce Service Cloud — Cases, Omni-Channel routing, Knowledge — and built the custom objects to represent ONDC orders, sellers and fulfillment status.
Order Tracker & AI Rollout
Built and connected the live order-tracker integration, then rolled out the AI-based chat agent across the DigiHaat app and WhatsApp, routing unresolved queries into Salesforce as Cases.
Run & Iterate
Ongoing support tuning routing rules, chat-agent responses and the order-tracker integration as DigiHaat's seller network and order volume grow.
What changed for DigiHaat (ONDC)
Real numbers only — pending metrics await client-approved data.
CSD order-issue resolution time, before vs. after
support channels unified — DigiHaat app chat and WhatsApp
order status on the Case, replacing manual cross-checking
categories on one order-tracker view
What used to be a multi-day back-and-forth to resolve an order complaint now closes in about five minutes — agents work from the order status already visible on the Case, and the AI chat agent on the app and WhatsApp handles routine “where is my order” questions before they reach an agent at all.
“Integrating our system with Salesforce was a great decision — with its help, we've been able to resolve customer issues in very little time.”
CSD Department Head
Customer Service Desk · DigiHaat
Salesforce Service Cloud — common questions
Is your support team stuck reconciling order status across systems too?
We build Salesforce Service Cloud desks — plus the integrations and AI layers around them — that give agents one accurate view instead of five open tabs.
Discuss a Service Cloud DeskRelated case studies
Explore our AI-Enhanced Salesforce Solutions, Salesforce development company in Pune, or start with a scoped AI Sprint.