notes✳
THE ENGINEERING FIELDGUIDEVOL. 001 / BACKEND & SYSTEM DESIGN

Beyond the code.
Into the system.

Backend and system design for SDE interviews and real jobs.
Understand what happens underneath, then design it, ship it, and recover it.

Try the design studioExplore the path
13 deep notesInteractive modelsA portfolio project
CASE STUDY 02TRY THE MECHANISM ↓
REQUEST / K1Interactive model
ClientIntent
APIContract
DatabaseDurable result

A client submits one job with intent key K1.

1 / 5
↳ The response is not the operation.
Trace what happens after a write commits.
CODE IS THE BEGINNING.UNDERSTANDING IS THE WORK.
FREE SYSTEM DESIGN LAB · NO SIGN-UP

Don’t just read about it.
Watch it carry load.

Kill a server behind a load balancer. Burst past a rate limiter. Expire a hot cache key and watch the database spike. Add a node to a hash ring and count the keys that move.

LOAD BALANCER / 4 SERVERSLive simulation
Clients14 req/sLoadbalancerround-robinServer 10 in progress or queuedServer 20 in progress or queuedServer 30 in progress or queuedServer 40 in progress or queued
S1
S2
S3
S4
Served
0
Failed
0
p95 latency
0.00 s
Capacity
32 req/s

Try it: make Server 2 slow. Round robin keeps sending it a quarter of the traffic, so its queue fills and p95 latency climbs. Switch to least connections and watch the queue drain. Then kill a server: requests fail until the next health check removes it from the pool.

Read the lesson behind it
THE METHOD

Make every idea
earn its place.

A note opens the idea. A model reveals the mechanism. An exercise asks you to use it. A project connects it to the system.

01UnderstandRead the explanation. Follow the exact state change.A clear mental model02DesignMake the assumptions, contracts, and tradeoffs explicit.A decision you can defend03BuildImplement the boundary. Test a case you did not rehearse.A reproducible artifact04DeliverFollow the same artifact through required checks and release.A verified workflow05RecoverInvestigate a failure. Mitigate it. Verify the user journey.An operational runbook
YOUR FIRST FIELDGUIDE

One service.
The whole story.

HTTP, API contracts, idempotency, PostgreSQL transactions and migrations, caching, queues, Docker, CI/CD, telemetry, incidents, and restores: the topics system design interviews keep asking about, taught through one service you actually build.

See all 13 notes
MechanismWhat happens?
SystemWhy this design?
OperationHow does it fail?
THE WORK YOU KEEP

Leave with something
you can explain.

Try a working Studio case, the first project workbook, and 4 open notes. The full path and Pro tools are ₹79, paid once.

01 / SYSTEM DESIGN CASE STUDY

The diagram is only the beginning.

Capture a baseline. Change a cache, add capacity, or break a dependency. Explain the result and export the architecture with your decisions.

BASELINECHANGEEXPLAIN
Try the open Studio case
02 / ENGINEERING PROJECT WORKBOOK

Keep the evidence behind the answer.

Build a jobs service in ten milestones. Link test runs, review failure behavior, and practice the follow-up questions an interviewer could ask.

BUILDVERIFYDEFEND
Open the first workbook
See everything in Notes Pro