How the Middle East’s Fastest-Growing Last-Mile Courier Broke the 25-Waypoint Ceiling to Scale Daily Deliveries

AT A GLANCE

• 5,000–6,000 daily delivery segments
• 50–60 stops per driver, per shift
• ~105,000 monthly optimizable orders
• Under 5 seconds API response time per route

BACKGROUND

This company is a fast-scaling courier and logistics operator based in Saudi Arabia, handling last-mile delivery across multiple urban zones. Its operations span small parcels through to larger freight consignments, with drivers covering assigned geofenced zones and executing up to 50–60 deliveries per shift.

The company had been building out a digital operations platform — covering driver dispatch, mobile app navigation, and delivery management — and had reached a critical scaling point where its approach to route planning was becoming a bottleneck.

With 5,000 to 6,000 delivery segments per day and ambitions to expand across additional cities and zones, it needed a route optimization engine capable of growing with the business.

THE CHALLENGE

The company had been relying on a Waypoints API for route sequencing. While workable at a small scale, a hard limit of 23–25 waypoints per API call was incompatible with the operational reality of a driver making 50 to 60 deliveries per shift. The team was forced to either manually split routes or leave optimization incomplete — neither of which was acceptable as volumes grew.

Beyond the waypoint ceiling, the existing approach offered no support for the operational constraints that define real-world courier logistics: time windows for priority shipments, vehicle capacity limits, driver break schedules, skill-based vehicle-to-parcel matching, or dynamic re-optimization when new orders arrived mid-shift.

Key Pain Points

• Incumbent solution capped at 23–25 waypoints per call — far below the 50–60 stops per driver, per shift
• No constraint handling: no time windows, no capacity limits, no driver break scheduling
• No priority job management: high-priority morning shipments could not be guaranteed first delivery
• No re-optimization capability when new orders arrived after routes were already planned
• No vehicle skill matching — no way to ensure the right parcel type reached the right vehicle type
• Manual route splitting workarounds were slow, error-prone, and unscalable
• ~30% of delivery locations lacked geocoordinates, creating gaps in optimization coverage

THE NEXTBILLION.AI SOLUTION

The company integrated NextBillion.ai’s Route Optimization API as the core planning engine for its delivery operations — with the Directions API powering driver navigation in the mobile app.

Route Optimization API

• Handles up to 500 stops per optimization call — removing the hard ceiling that had blocked the company’s scaling, and comfortably supporting 50–120 stops per driver route.
• Time window constraints allow priority shipments to be assigned and sequenced first — ensuring high-value morning deliveries are always completed within the required window.
• Vehicle capacity parameters ensure each driver’s load respects physical limits, with multi-dimensional capacity support across units (weight, volume, number of items).
• Skills-based matching assigns parcels to appropriately equipped vehicles — supporting specialized delivery types such as refrigerated goods or items requiring lift equipment.
• Driver break scheduling factors in mandatory rest periods without disrupting the overall route sequence.
• Re-optimization capability enables new orders to be inserted into an already-planned route mid-shift without rebuilding the entire schedule from scratch.
• Geofence-aware input: the company passes only stops within a driver’s assigned zone boundary, keeping optimization problems well-scoped and response times consistently under 5 seconds.

Directions API

• Provides turn-by-turn navigation integrated directly into the driver mobile app, layered over existing map infrastructure.
• Latitude/longitude input accepted directly — compatible with the company’s existing geocoding data pipeline.

RESULTS

Waypoint ceiling eliminated — From a hard cap of 23–25 stops per API call to 500+ stops per NextBillion.ai optimization call — fully accommodating the 50–60 stop reality of every driver’s shift.

Priority delivery guarantee — Time window constraints ensure high-priority morning shipments are always sequenced and completed first, within their delivery window, with no unassigned jobs for geofenced zones.

Full constraint support — Vehicle capacity, driver breaks, service time, and skills-based matching are now embedded in every optimization — reducing misrouted deliveries and improving fleet utilization.

Dynamic re-optimization — New orders arriving mid-shift can be inserted and re-optimized without disrupting the existing route plan — removing a key operational bottleneck as order volumes grow.

~105,000 monthly orders optimized — Daily volumes of 5,000–6,000 segments scale to roughly 105,000 optimized orders per month across all driver zones.