Operations Platform for Last-Mile Delivery Company
78% → 91%
Delivery success rate improved to 91%
2 hours → 15 min
Dispatch time reduced to 15 minutes per batch
60%
Reduction in WISMO customer calls
Industry
Logistics & Supply Chain
Services
Digital Platform Engineering, Mobile App Development
Tech Stack
React Native (Offline-capable), Next.js, NestJS, PostgreSQL
Duration
4 months
What Needed to Change
A last-mile delivery company managing 500 delivery partners across multiple zones was running its entire operation through WhatsApp groups and Google Sheets.
Morning dispatch — assigning delivery batches to partners — was a 2-hour manual process of matching order lists to partner availability, zone coverage, and vehicle capacity, all done in a spreadsheet and communicated via broadcast messages. Proof of delivery was paper-based — riders carried printed sheets, collected customer signatures, and returned them to the warehouse at day-end, meaning delivery confirmations were only available 8–12 hours after actual completion. Customer inquiries about delivery status (WISMO — Where Is My Order) flooded the support team because there was no tracking visibility for end customers. Management had no real-time data on delivery success rates, partner performance, or issue resolution — weekly performance reviews were based on manually compiled figures that were already a week old.
How We Solved It
Brihat Infotech built a three-sided operations platform covering admin dispatch management, delivery partner mobile operations, and customer tracking visibility. The Admin Dashboard automated batch creation — orders were ingested from the client's order management system via API, clustered by delivery zone and date, and assigned to available partners based on capacity, zone familiarity, and performance score — reducing the 2-hour manual dispatch to a 15-minute supervised process. The Delivery Partner Mobile App (React Native) supported both online and offline modes — riders in low-signal areas could operate fully offline, with data syncing automatically on reconnection. Features included route navigation via Google Maps integration, digital proof of delivery with photo capture and GPS-stamped confirmation, cash collection tracking, and in-app communication with the operations team. A Customer Tracking Page provided live status updates — dispatched, out for delivery, delivered, or exception — eliminating the need to call customer support for order updates. A Partner Performance Dashboard tracked delivery success rates, on-time rates, customer satisfaction scores, and average deliveries per day per partner — enabling data-driven performance management and incentive allocation.
What We Built
Offline-first React Native app for delivery partners — order assignment, navigation, and POD confirmation work without connectivity
GPS-stamped, timestamped photo proof-of-delivery with mandatory geo-verification
Zone-based auto-assignment algorithm dispatching 300–400 orders per shift in under 15 minutes
Live dispatch console: real-time partner map, order status, battery levels, and SLA breach alerts
Customer live-tracking link via SMS with real-time partner location and ETA on a branded page
Partner performance analytics: delivery success rates, failure reasons, and route efficiency by zone
Why This Stack?
React Native was the only viable choice for offline-first delivery apps at this scale — Flutter's offline sync story was less mature at the time. We used SQLite on-device with a custom sync queue that replays failed operations in order when connectivity returns. Redis powers the real-time dispatch console with sub-100ms partner location updates pushed via WebSockets.
The Results
78% → 91%
Delivery success rate improved to 91%
2 hours → 15 min
Dispatch time reduced to 15 minutes per batch
60%
Reduction in WISMO customer calls
100%
Automated partner performance tracking
The client scaled from a single city to four metro areas within eight months of launch, using the same platform with only zone configuration changes. Delivery disputes dropped sharply once geo-verified photo POD was mandatory. The dispatch team went from managing order assignment manually to overseeing the algorithm — a shift from execution to exception management.
Lessons We're Taking Forward
Offline-first is not a feature — it's an architectural commitment
Adding offline support to an online-first app requires rethinking every assumption about data freshness, conflict resolution, and sync ordering. We designed the SQLite on-device schema and sync queue before writing any UI — retrofitting it would have been a full rewrite.
GPS verification kills fraudulent proof-of-delivery at scale
Before geo-stamped POD, ~4% of deliveries had disputed status between partners and customers. After mandatory geo-verification on photo capture, false delivery claims dropped to under 0.3% in the first month — the client recovered more in avoided re-deliveries than the platform cost.
Auto-assignment algorithms need a human override, always
Our zone-based dispatch algorithm handles 95% of assignments optimally. The remaining 5% involves edge cases — a partner breaking down, an order requiring special handling — that no algorithm will anticipate. A one-click manual override kept operations flowing instead of creating dispatcher-vs-algorithm conflicts.
Related Case Studies
Building something similar?
We've solved problems like this platform before. Let's talk about yours.