Venture 01
CarryIt — the delivery layer that keeps money in the neighbourhood.
Every day thousands of local trips happen with empty capacity, while merchants hand 25-30% of an order to a national platform. CarryIt matches those two facts.
The gap
National delivery apps price out independents and charge customers twice — once on fees, once on inflated menu prices.
The insight
Local delivery is a routing problem, not a fleet problem. AI can extract enough efficiency to halve the take rate.
The wedge
Merchants who already deliver themselves, badly. They convert fastest and bring their own order volume.
Strategy
Development roadmap
Phase 1 — One corridor
Launch in a single dense corridor with 30-50 merchants and a closed courier pool. Prove sub-40-minute median delivery and a 12-15% take rate that still beats incumbents for the merchant.
Phase 2 — Density flywheel
Batch orders with AI routing so one courier trip serves 2-4 drops. Every point of batching efficiency goes back into courier pay and lower merchant fees, not into margin.
Phase 3 — Merchant OS
Add inventory sync, repeat-customer messaging and demand forecasting. CarryIt becomes the operating layer for the merchant, and delivery becomes the retention hook.
Phase 4 — City-to-city replication
Package the launch playbook: corridor selection model, courier acquisition funnel, merchant onboarding in under 48 hours. Repeat per metro with a two-person local team.
Technology
Where the AI actually earns its keep
Routing & batching
Predicts drop clusters and matches them to courier trip patterns instead of dispatching one job at a time.
Demand forecasting
Tells merchants what to prep and when, and tells couriers when to be available.
Trust scoring
Blends delivery evidence, timing and customer signal into a live reliability score per courier.
Pricing engine
Dynamic but capped pricing that protects the customer from surge shock while keeping courier pay above local floor.