# Design Food Delivery System like Uber Eats (Mock Interview)

Aced (formerly Exponent) publishes a public search API over its indexed pages; the guide is at https://www.aced.io/agents/agent-access.

Source: https://www.aced.io/blog/design-food-delivery-uber-eats-system-design-interview
Source observed: 2026-11-05T00:17:09.318Z
Last checked: 2026-11-05T00:17:09.318+00:00
Indexed: 2026-11-06T01:23:09.236+00:00

---

# Design Food Delivery System like Uber Eats (Mock Interview)

**System Design** · Omar Hijazi · Last updated a month ago

We use this mock interview to explain how Neeraj, an Engineering Manager at eBay, approached designing Uber Eats, a food-delivery app. For deeper preparation, see our [System Design Interview Course](https://www.aced.io/courses/system-design-interviews?src=blog-top-callout&ref=exponent-blog.ghost.io). We also link sample answers for [Instagram](https://www.aced.io/courses/system-design-interviews/mock-interview-design-instagram?src=blog&ref=exponent-blog.ghost.io), [TinyURL](https://www.aced.io/courses/system-design-interviews/url-shortener?src=blog&ref=exponent-blog.ghost.io), and [YouTube](https://www.aced.io/courses/system-design-interviews/design-youtube?src=blog&ref=exponent-blog.ghost.io).

## Functional requirements

A food-delivery system includes restaurants, customers, delivery drivers, and orders. Restaurants should be able to add, update, and remove menu items and related information. Customers should be able to view, search, and filter restaurants by distance, food type, and estimated delivery time.

## Non-functional requirements

The main system attributes are scalability, data consistency, security, availability, and operational latency. Menu and restaurant updates can take a few seconds to propagate, but high availability is a top priority: slightly outdated data is preferable to a platform that crashes.

Security should prevent users from accessing or manipulating information they do not own. Identity access management and logging mechanisms are possible controls. Desired response targets are **under 150 milliseconds** for viewing pages and **under 400 milliseconds** for restaurant searches. Image uploads and new-menu-item uploads can tolerate somewhat higher latency.

## Understanding the scale of our data

For this example, we assume that we add **100 restaurants** and **10,000 customers daily**. Over 20 years, that produces about **~1.73 million restaurants** and **~73 million customers**. We also estimate about **100 million restaurant views and searches per day**.

## Defining data models

A data model lists each component’s fields and its relationships with other components.

- **Menu item:** item name, currency, price, and image.
- **Restaurant:** name, unique ID, address, city and location, and distance from the user’s location.
- **Customer:** name, delivery address, country, latitude, and longitude.

Latitude and longitude can be used to calculate distance.

## Calculating distances

We can calculate delivery times between restaurants and customers with **GeoHashing**. GeoHashing divides the world into grids and represents each location with a short alphanumeric string. Longer strings identify smaller, more specific areas; shorter strings identify larger areas.

Grouping restaurants in a small area, such as a shopping center, lets us treat them as one data unit when delivery time is approximately constant across the group. This simplifies the design.

## Database design

A relational database could store customers, menu items, restaurants, and other entities in separate tables. Multiple read replicas could improve read performance; a read replica is a copy of the primary database dedicated to read requests.

Because databases typically receive more reads than writes, replicas are useful. At the example’s high scale—described as **~173 million users** in the article—a NoSQL database such as Cassandra may be more appropriate because it supports automatic sharding. Sharding distributes a database across machines to increase read/write throughput, storage capacity, and availability.

## Designing specific services

We can begin with a layer that orchestrates services across iOS, Android, mobile web, and desktop web so back-end resources can be shared. A restaurant service can sit behind a load balancer that distributes traffic across multiple machines.

### Handling images

Menu images can be expensive to send over the network and store in a database, so object storage such as S3 is a more efficient option. An image service can apply a moderation policy using machine-learning image-classification models to check accuracy and quality.

We can evaluate the model on a **sample data set** using precision and recall:

- **Recall:** the proportion of relevant items found. The article’s **sample fishing analogy** catches 80 of 100 fish, giving 80% recall.
- **Precision:** the proportion of retrieved items that are relevant. In that same **sample analogy**, 80 fish and 80 rocks produce 50% precision.
- A smaller **sample net** that catches 20 fish and no rocks gives 20% recall and 100% precision.

If validation is slow, we can optimize the API or add resources such as GPUs or multiple CPUs. We can also speed retrieval with caching, using LRU (Least Recently Used) or compression.

### Search functionality

A search service identifies nearby restaurants by location. Elasticsearch supports geo-queries: store restaurant coordinates in Elasticsearch and query with the customer’s coordinates.

To filter by delivery time, we can create a polygon map describing the region each restaurant can serve within a specified period and store it in a database such as DynamoDB. Customers can then see restaurants able to deliver within their requested time.

Elasticsearch uses a block k-dimensional tree, or **BKD**, described here as an improved geo-hash version.

## Learn everything you need to ace your system design interviews

We provide free interview guides, insider tips, and courses through Aced. [Create your free account](https://www.aced.io/signup?src=blog).

## Related blog posts

- [System Design Interview Prep & Questions (2026 Guide)](https://www.aced.io/blog/system-design-interview-guide) — 17 min read
- [System Design Interview: Design a Rate Limiter](https://www.aced.io/blog/rate-limiter-system-design) — 8 min read
- [The 10 Best System Design Books to Sharpen Your Skills](https://www.aced.io/blog/best-system-design-books) — 9 min read
- [How to Whiteboard for System Design Interviews](https://www.aced.io/blog/how-to-whiteboard-for-system-design-interviews) — 8 min read

---

[Read this page’s summary](https://www.aced.io/agents/api/public/v1/page-tldr?url=https%3A%2F%2Fwww.aced.io%2Fblog%2Fdesign-food-delivery-uber-eats-system-design-interview&trace=t_679ff9ab98f3)

## Ask a follow-up about this page

Need a more specific answer? Ask the exact question you have. It returns a fresh, cited answer from Aced (formerly Exponent)’s knowledge base. No account or API key required.

Open this URL with your own question, URL-encoded. Include your question in q. All research inputs are optional; no account or setup.

`GET https://www.aced.io/agents/agent/faq?q={question}&trace=t_679ff9ab98f3`

[Tool guide and examples](https://www.aced.io/agents/agent-access) · [OpenAPI](https://www.aced.io/agents/api/public/v1/openapi.json) · [Page TLDR form](https://www.aced.io/agents/agent-tools/page-tldr?url=https%3A%2F%2Fwww.aced.io%2Fblog%2Fdesign-food-delivery-uber-eats-system-design-interview). Opening a form does not execute a tool.