System Design Cheat Sheets

Offline, self-contained HTML reference guides for system design and event-driven architecture interview prep. Open any file directly in a browser — no build step or server required.

Start here: cheatSheet/index.html

View on GitHub: Clicking .html in the repo shows source code, not a rendered page (GitHub serves HTML as plain text). Use:

What you want	Open this
Browse in GitHub now	cheatSheet/github/v15/index.md — 40 systems (Markdown)
HTML in browser (no clone)	html-preview or see cheatSheet/VIEW_ON_GITHUB.md
Full interactivity	Enable GitHub Pages (Settings → Pages → GitHub Actions) → https://eddyclhung.github.io/systemDesign/

Pre-rendered HTML (*_github.html) is for GitHub Pages or preview tools — not the repo file click-to-preview.

flowchart TB
    subgraph repo["systemDesign"]
        README["README.md"]
        INDEX["cheatSheet/index.html"]
        subgraph cheatSheet["cheatSheet/"]
            v10["SystemDesign_Complete_v10.html"]
            v15["system_design_cheatsheet_v14.html (v15)"]
            eda["EDA_Interview_Question_Bank_Complete.html"]
        end
    end

    INDEX --> v10
    INDEX --> v15
    INDEX --> eda
    INDEX --> ai

    v10 --> bbg["ByteByteGo Reference"]
    v10 --> cards["System Cards"]

    bbg --> foundations["Foundations ×3"]
    bbg --> vol1["Vol. 1 ×11"]
    bbg --> vol2["Vol. 2 ×11"]
    bbg --> appendix["Cloud appendix · CLI commands"]

    cards --> c26["26 systems"]

    v15 --> framework["Delivery Framework + worksheet"]
    v15 --> dbchooser["DB Chooser"]
    v15 --> quickref["Quick Reference + Cloud cmds"]
    v15 --> s40["40 systems · 4E / 18M / 18H"]

    eda --> edaq["26 EDA questions · 78 follow-ups · 12 diagrams"]
    ai["AI_Interview_Prep.html"] --> aiq["RAG · Agents · Evals · Player Service"]

    goal{{"What's your goal?"}}
    goal -->|"Study ByteByteGo"| v10
    goal -->|"Practice interviews"| v15
    goal -->|"Drill Kafka / EDA"| eda
    goal -->|"AI eval / take-home"| ai

Loading

Files

File	Description
cheatSheet/index.html	Landing page — pick v10, v15, or EDA
cheatSheet/EDA_Interview_Question_Bank_Complete.html	EDA interview bank — 26 questions, follow-ups, diagrams
cheatSheet/AI_Interview_Prep.html	AI interview prep — RAG, agents, hallucination handling (Player Service)
cheatSheet/AI_TOPICS_INTERVIEW_PREP.md	AI prep source markdown
cheatSheet/github/v15/index.md	v15 on GitHub — Markdown (renders in repo browser)
cheatSheet/system_design_cheatsheet_v15_github.html	v15 HTML — pre-rendered for Pages / html-preview
cheatSheet/VIEW_ON_GITHUB.md	Viewing guide — what works where
cheatSheet/SystemDesign_Complete_v10_github.html	v10 for GitHub — pre-rendered system cards
cheatSheet/github/	Markdown per system
cheatSheet/SystemDesign_Complete_v10.html	ByteByteGo Vol. 1 & 2 deep-dive + 26 system cards + cloud appendix
cheatSheet/system_design_cheatsheet_v14.html	Staff+ interview prep (v15 content) — 40 systems
CHANGELOG.md	Version history
scripts/validate_systems.py	Card completeness validator

Quick start

open cheatSheet/index.html
# or directly:
open cheatSheet/system_design_cheatsheet_v14.html
open cheatSheet/SystemDesign_Complete_v10.html
open cheatSheet/EDA_Interview_Question_Bank_Complete.html
open cheatSheet/AI_Interview_Prep.html

Validate card data after edits:

python3 scripts/validate_systems.py
python3 scripts/build_github_view.py      # regenerate Markdown + Pages landing HTML
python3 scripts/build_prerendered_html.py  # regenerate *_github.html
python3 scripts/build_quick_fire_html.py   # regenerate interview-quick-fire.html
python3 scripts/enrich_quick_fire_ladder.py  # Weak/Strong/Staff+ on quick-fire patterns

Study paths

Staff answer ladder (daily drill)

Every quick-fire pattern and v15 Deep dive uses three rungs. Practice climbing one rung per follow-up — never jump straight to Staff+.

Rung	Say this	Signals
🔴 Weak	Tool name only	Pattern recall
🟡 Strong	Pattern + why it fits this workload	Credible design
🟢 Staff+	Failure mode + metric + revisit trigger	Operated production

15 min/day: Pick 3 patterns from interview-quick-fire.md — say Weak out loud, then Strong, then Staff+ with one metric. 45 min mock: One v15 card — deliver Strong on architecture, Staff+ on each deep dive when probed.

python3 scripts/enrich_quick_fire_ladder.py   # after editing patterns
python3 scripts/build_quick_fire_html.py      # colorful HTML
python3 scripts/build_github_view.py          # v15 MD deep dives

4-week interview prep (v15)

Week	Focus	Systems
1	Easy + framework	Bitly, Dropbox, GoPuff, Google News — memorize delivery framework + estimation worksheet
2	Medium reads/writes	WhatsApp, News Feed, Yelp, Rate limiter, Notification, Autocomplete — practice 45-min mocks
3	Hard distributed	Uber, YouTube, Payment, Kafka, KV store, Google Docs — deep dives + failure modes; pair Kafka card with EDA question bank
4	Gap fill + review	Maps, Email, S3, Wallet + any weak cards — interview mode + print for flashcards

ByteByteGo alignment (v10 → v15)

Use v10 chapters for theory, then drill the matching v15 card:

• URL Shortener → Bitly
• Notification System → Notification system (APNs/FCM)
• Google Maps → Google Maps
• Object Storage (S3) → S3 object storage
• Distributed Email → Distributed email (Gmail)
• Payment System → Payment system + Digital wallet
• Distributed Message Queue (Kafka) → Message queue (Kafka) + EDA question bank

system_design_cheatsheet_v14.html (v15)

Staff+ prep built around a repeatable delivery framework.

Delivery framework — six-step flow plus clarifying questions, latency/QPS numbers, and a live estimation worksheet.

Architecture framework — universal skeleton, two decision questions (read vs write · money vs likes), five archetypes with blueprints, and a live Splitwise derivation. Based on Firdaus Jawed’s framework (Javarevisited).

Database chooser — comparison table (open by default), decision tree, per-DB deep dives, anti-patterns.

Quick reference — CAP, protocols, consistency patterns, cloud services + AWS commands (full tables in v10).

Interview quick-fire — cheatSheet/interview-quick-fire.md — 50 patterns with Weak → Strong → Staff+ ladders, DMOP deep-dive spine, and severity callouts. Colorful view: interview-quick-fire.html.

40 systems by difficulty:

Easy (4)	Medium (18)	Hard (18)
Bitly	Ticketmaster	Instagram
Dropbox	WhatsApp	YouTube Top K
Local delivery (GoPuff)	FB News Feed	Uber
News aggregator (Google News)	Tinder	Robinhood
	LeetCode	Google Docs
	Distributed rate limiter	Distributed cache
	FB Live Comments	YouTube
	FB Post Search	Web crawler
	Yelp	Ad click aggregator
	Strava	Job scheduler (Airflow)
	Online auction (eBay)	Payment system (Stripe)
	Price tracking	Metrics monitoring (Datadog)
	Notification system (APNs/FCM)	Message queue (Kafka)
	Search autocomplete (Google)	Distributed key-value store
	Unique ID generator (Snowflake)	Nearby friends
	Hotel reservation (Booking.com)	Google Maps
	Gaming leaderboard	Distributed email (Gmail)
	S3 object storage	Digital wallet (Apple Pay)

UX features

• Sidebar + topbar search
• Interview mode (5 essential tabs)
• Keyboard shortcuts (/, j/k, 1–5, Esc)
• Deep links (#card-12, #uber-ride-sharing)
• Theme, tabs, studied progress — persisted in localStorage
• Prev/Next card navigation
• Copy script button
• Print stylesheet

SystemDesign_Complete_v10.html

Two-tab cheatsheet: ByteByteGo reference + 26 system cards.

Quick Reference appendix — database selection, CAP, protocols, AWS/GCP/Azure with CLI commands and Java single-server equivalents.

Features: topic search, chapter sidebar, print layout. Links to v15 Staff+ prep in header.

EDA_Interview_Question_Bank_Complete.html

Focused event-driven architecture drill bank — complements v15 cards like Kafka, message queue, and notification systems.

26 questions across 10 categories: delivery semantics, ordering, failure handling, Kafka internals, patterns, backpressure, schema evolution, observability, async vs sync, and real-world failures.

Each question includes a strong-hire answer, a trade-off to deliver proactively, expandable follow-up Q&A (78 total), and inline SVG diagrams (12 across categories).

UX features

• Category sidebar + search
• Mark reviewed + progress bar
• Hide reviewed / hard-only filters
• Expand / collapse all
• Keyboard shortcuts (/ search, h hard only, r hide reviewed, Esc clear search)

Open locally or via GitHub Pages — fully self-contained except Google Fonts (loads when online).

Which one to use?

• Studying ByteByteGo or need cloud CLI reference → SystemDesign_Complete_v10.html
• Practicing live interviews with framework, DB chooser, scripts → system_design_cheatsheet_v14.html (v15)
• Drilling Kafka / delivery semantics / EDA follow-ups → EDA_Interview_Question_Bank_Complete.html

All files are standalone — styles and scripts embedded inline. Optional: extract system data with python3 scripts/extract_systems.py.