Uber System Design - Driver finding use case 👇

Most candidates say:
"It finds the nearest driver" ❌
 
That's not wrong. But it's incomplete.
Here's what's actually happening 👇
 
Step 1 → Geohashing

Imagine the entire world map
broken into tiny hexagonal cells.
Each cell has a unique ID (a hash).
 
When you open Uber —
your phone's coordinates get converted into one of these cell IDs instantly.
 
Uber now ONLY looks for drivers
inside that cell + its neighboring cells.
 
Instead of scanning 5 million drivers globally,
it scans maybe 50-100 drivers nearby.
 
That's why matching is fast.
Not magic. Just smart filtering. ⚡
 
Step 2 → WebSockets
 
Uber uses WebSockets --- one permanent open connection between driver's phone and Uber's server.
 
Every 4 seconds, driver location
is pushed automatically.
No asking. No waiting.
 
That's how the map moves smoothly
in real time on your screen. 📡
Step 3 → ETA ≠ Distance
This is the part most people miss.
 
Uber doesn't assign the closest driver. It assigns the FASTEST driver.
Uber calculates ETA using:
→ Real-time traffic data
→ Road type (highway vs lane)
→ Turn restrictions & one-ways
→ Historical speed patterns on that route
 
All of this computed per driver,
in parallel, in milliseconds. 🧠
 
 
This exact question was asked at:
→ Google
→ Swiggy
→ Uber itself 😅
 
Save this 📌 before your next interview.
 
Follow @CompileCodeClub —
I break down how real apps work,
one concept at a time.
 
#systemdesign #uber #interviewprep #compilecodeclub #geohashing
 
[systemdesign, uber, geohashing, websockets, softwareengineering, interviewprep, techjobs, sdeinterview, coding, developer, programmer, backenddevelopment, 100daysofcode, codingtips, reels, instareels, viralreels, educationalreels, learnontiktok, techreels, indiandeveloper, indiantechcommunity, placements, campusplacement, interviewquestions, googleinterview, crackingthecodinginterview, webdevelopment, techcommunity, compilecodeclub]

Видео Uber System Design - Driver finding use case 👇 канала Compile Code Club