The archive · Logistics & Supply · Product decision · 2022-2026
Dayjob bets AI agents on existing ERPs can replace manual waste-truck scheduling
Ex-Deliveroo founder pivots from waste ERP to AI scheduling agents: 60-second route plans, 8%+ fleet gains, $496K ARR at YC launch.
Dayjob
What the business is
AI scheduling agents for short-haul trucking that plug into a fleet's existing ERP and continuously re-optimise routes, driver changes, and exceptions in real time.
How it started
George Postlethwaite (an early Deliveroo employee who had built Gaea, an ERP for e-waste recyclers) and Fred Fooks, who met at Oxford, started building ERP and inventory software for waste operators in October 2022 after watching planners at depots manually drag jobs onto routes each morning; they generated $150K in revenue.
What happened
Two lessons emerged: large operators would not migrate ERPs, and better UI did not fix the real problem - manual decision-making at scale. In January 2025 they pivoted to Dayjob, layering AI scheduling agents on top of existing ERPs instead of replacing them. Sales began August 2025; by launch ARR was $496K and grew to $653K within seven months. Customers in waste management report 8%+ efficiency gains from day one and 50% better hit rates on customer time slots; one customer cited an £800K first-year revenue uplift. Dayjob joined YC's Spring 2026 batch as a five-person London team.
How it ended up
Still live. Dayjob is expanding from UK waste fleets into the US short-haul market, with three more agents covering the full operational workflow planned later in 2026.
Background
Dayjob is a London startup building autonomous AI agents for industrial logistics, starting with waste and short-haul trucking. Its scheduling agent plugs into a fleet's existing ERP and rebuilds the full route plan in about 60 seconds - a task human transport planners spend 60 to 90 minutes on each morning and that is usually obsolete within an hour.
Founders George Postlethwaite (an early Deliveroo employee and builder of Gaea, an ERP for e-waste recyclers) and Fred Fooks met at Oxford. They began with ERP software for waste operators in October 2022 and generated $150K in revenue, but learned operators would not migrate ERPs and that better UI did not solve manual decision-making. In January 2025 they pivoted to Dayjob, layering AI agents on existing systems.
By the YC Spring 2026 batch, Dayjob reported $496K ARR, 8%+ efficiency gains for waste customers from day one, 50% better on-time slot performance, and one customer citing an £800K first-year revenue uplift; ARR reached $653K within seven months of first sales.
The bet is that scheduling is a cognitive bottleneck - planners burn out within about 18 months and cannot be hired fast enough - so an agent that continuously re-optimises routes, driver hours, and live exceptions becomes the operating layer of industrial logistics.
What has to be true
- Route planning looks like a feature problem but is a decision-making bottleneck: humans cannot continuously re-optimise hundreds of live constraints, and planners burn out within about 18 months.
- Waste was a deliberate wedge: its high variability and tight regulatory rules make it the hardest planning environment, so success there is proof the agent generalises.
- Layering on existing ERPs removed the adoption barrier - operators refused to migrate - turning Dayjob from a replacement sale into an outcome sale.
What can be applied
Don't fight incumbents' installed base: operators would not migrate ERPs, so Dayjob layered agents on top and sold the outcome - plans that stay current - instead of a replacement system.
Aftermath
As of September 2026, Dayjob remains active as a five-person London team backed by Y Combinator's Spring 2026 batch. It is live with fleets in the UK and launching in the US short-haul market, reporting $653K ARR within seven months of first sales. The company plans three more agents covering pricing, booking, and the rest of the operational workflow, and frames success as becoming the layer that runs transport operations in real time.
Sources
- Dayjob: AI Agents for Industrial Logistics (Launch YC)
- Dayjob: AI Scheduling for Short Haul Trucks
- YC-Backed Dayjob Launches AI Agent for Waste Transport Scheduling
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card