← Back to selected work

Munch:ON · Acquired by Careem

Real-Time Delivery Operations Platform

Operations tooling for a batched corporate meal-delivery model: planning company orders, assigning delivery runs, monitoring drivers, and resolving real-time exceptions.

Acquisition announced in June 2022 · Read announcement

Role
Frontend engineer building high-density operations interfaces, nested order tables, map workflows, and real-time driver visibility.
System scope
Advance-order batching, company-level delivery orders, expandable customer-order detail, map-based assignment, and driver/order analysis.
Scale & impact
Supported systems used to manage approximately 3,000 delivery orders per day. Contributed to dashboard work associated with 9% month-over-month growth and approximately $8K in monthly savings.

Problem

Munch:ON operated a batched corporate meal-delivery model, where advance orders were grouped by company and coordinated across delivery runs. Operations teams needed to inspect company-level orders, drill into individual customer requests, assign delivery work, and monitor active drivers without losing context or slowing down under dense data.

What I built

  • Built a high-performance 17-column company-order table with expandable customer-order subrows.
  • Built a Mapbox dispatch map for drag-and-drop order assignment and driver/order analysis.
  • Added operational stats cards with a controlled 10-minute refresh cadence for planning and monitoring.
  • Contributed to the V2 workflow for delivery planning and map-based dispatch.

Constraints

  • Company delivery orders could contain multiple individual customer orders, creating multi-level data that needed to remain easy to scan.
  • Operational users needed both planning workflows and active delivery visibility in the same system.
  • Assignment workflows required spatial context, clear status signals, and drag-and-drop interaction.
  • Internal tools prioritised speed, clarity, and reliability over decorative UI.

Technical decisions

  • Applied table and list virtualization to maintain responsiveness in dense views.
  • Designed expandable nested-table interactions so operators could access customer-level detail without losing company-order context.
  • Used map-based spatial interaction for delivery assignment and live driver/order analysis.
  • Kept summary metrics on a slower refresh cadence while operational order and driver views remained current.

Outcomes

  • Supported systems used to manage approximately 3,000 delivery orders per day.
  • Contributed to dashboard work associated with 9% month-over-month growth.
  • Contributed to operational efficiency associated with approximately $8K in monthly savings.
  • Improved the operational team’s ability to inspect batch-level and customer-level delivery data in one workflow.

Technologies

Next.jsReactReact VirtualizedReact Tablereact-map-glMapboxNode.js
Discuss a similar product