System Design AI

Designs

Pick a system. Work through the problem. Compare your approach.

Create your own

All designs 148

Ordered by design number

Company tags are community-reported. Counts on cards show how many people reported that design.

121. Design an LLM Evaluation Platform
GPU batch generation behind a shared cache, sandboxed pass@k, calibrated judges, paired statistics, release gates, contamination checks.
ClassicMedium
Generative AI AI infrastructure AWS
+3
Evaluation LLMs MLOps
Worked design
Practice
122. Design an Experiment Tracking and Model Registry Service
Experiment tracking and a model registry: non-blocking logging, chunked metric curves, content-addressed artifacts, lineage and a promotion gate. One-hour boards for junior, senior and staff, with the theory behind them.
ClassicMedium
Machine learning AI infrastructure AWS
+3
Experiment tracking MLOps Model registry
Worked design
Practice
123. Design a Hyperparameter Tuning Service
A tuning service like Vizier: random and Bayesian search, ASHA and population-based training, trials on a shared GPU cluster.
ClassicMedium
AI infrastructure AWS Hyperparameter tuning
+2
MLOps Optimization
Worked design
Practice
124. Design Multi-Tenant LoRA Adapter Serving
Serve 10,000 LoRA adapters for 2,000 tenants on one shared base model, with mixed-adapter batches, tiered adapter caches, affinity routing and base upgrades.
ClassicMedium
Machine learning Generative AI AI infrastructure
+4
AWS Inference LLMs LoRA
Worked design
Practice
125. Design a Prompt and Semantic Cache for LLMs
Exact, semantic and prefix caching in front of LLMs: scoped keys, distances with an error budget, tenant isolation, invalidation and measured savings.
ClassicMedium
Caching Generative AI AI infrastructure
+4
AWS Embeddings LLMs Retrieval
Worked design
Practice
126. Design Model Distillation at Scale
A 70B teacher distilled into an 8B student: transfer sets, logit and on-policy distillation, per-slice gates, escalation routing, a refresh loop.
ClassicMedium
Machine learning Generative AI AWS
+3
Distillation LLMs Training
Worked design
Practice
127. Design a Model Quantization and Compression Pipeline
Every new LLM version shrunk to FP8, INT8 or INT4 with a smaller KV cache, gated against its BF16 parent and benchmarked on the serving GPU.
ClassicMedium
Machine learning Generative AI AWS
+3
Inference LLMs Quantization
Worked design
Practice
128. Design a Synthetic Data Generation Pipeline
Teacher models write training data for smaller students: conditioned and evolved prompts, verified answers, dedup, decontamination, versioned datasets with lineage.
ClassicMedium
Machine learning Generative AI AWS
+3
LLMs Synthetic data Training
Worked design
Practice
129. Design an Image-Text Training Data Pipeline
Billions of image-text pairs from Common Crawl: polite fetching, CLIP scoring, dedup, recaptioning, WebDataset shards and takedowns.
ClassicMedium
AI infrastructure AWS Computer vision
+2
Data pipeline Multimodal
Worked design
Practice
130. Design a Visual Search System
Search by photo over a billion images: contrastive embeddings from engagement pairs, object crops, IVF-PQ with re-scoring, versioned indexes.
ClassicMedium
Machine learning AWS Computer vision
+3
Embeddings Retrieval Visual search
Worked design
Practice
131. Design Retrieval-Augmented Generation
Answers from private documents with citations: chunking, BM25 + vector retrieval, reranking, permission filters, grounding checks and evaluation.
ClassicMedium
Generative AI AWS LLMs
+2
RAG Retrieval
Worked design
Practice
132. Design Gmail Smart Compose
Next words within 100 ms of a keystroke: conditional language modelling, a distilled student on draft-pinned KV caches, private n-grams, DP and canary audits.
ClassicMedium
Machine learning Generative AI Autocomplete
+3
AWS Latency LLMs
Worked design
Practice
133. Design Google Translate
Translation for 20 to 240 languages: one multilingual encoder-decoder, mined and back-translated data, distilled students, COMET and MQM, quality-aware serving: one-hour boards for junior, senior and staff, with the theory.
ClassicMedium
Machine learning Generative AI Multilingual
+2
Sequence-to-sequence Translation
Worked design
Practice
134. Design Text-to-Image Generation
Prompt to four images in seconds: latent diffusion, guidance, a distilled few-step student, recaptioned data, safety on both sides, watermarks and C2PA.
ClassicMedium
Generative AI AWS Diffusion
+2
Image generation Multimodal
Worked design
Practice
135. Design Image Captioning
Captions read aloud as alt text: encoder, bridge and decoder, CLIP-filtered data, beam search, hallucination-aware evaluation, lanes on one GPU pool.
ClassicMedium
Machine learning Generative AI Captioning
+2
Computer vision Multimodal
Worked design
Practice
136. Design Realistic Face Generation
Faces of people who do not exist: a style-based GAN, truncation, latent-space edits, selfie inversion, provenance and memorisation audits.
ClassicMedium
Generative AI AWS Diffusion
+2
GANs Image generation
Worked design
Practice
137. Design Ad Click Prediction
Predict the chance a person clicks an ad, calibrated for the auction, at 10 B requests a day: sampling and its correction, DCN-V2, delayed clicks.
ClassicMedium
Machine learning Ads AWS
+2
Click-through prediction Ranking
Worked design
Practice
138. Design Harmful Content Detection
Find and act on posts that break the rules in text, images, video and live: hash matching, a calibrated multimodal model, review ranked by expected harm, prevalence: one-hour boards for junior, senior and staff, with the theory.
ClassicMedium
Machine learning AWS Classification
+2
Multimodal Trust and safety
Worked design
Practice
139. Design Similar Listings for a Vacation Rental Platform
Pick 12 homes a guest could book instead, from listing embeddings learned on browsing sessions, filtered by dates and party size, then ranked.
ClassicMedium
Machine learning AWS Embeddings
+3
Marketplace Recommendations Retrieval
Worked design
Practice
140. Design People You May Know
Suggest people a member knows out of a billion: bounded friends of friends, affiliations and contacts, two ranking heads, GNN embeddings, privacy and abuse.
ClassicMedium
Machine learning AWS Graph ML
+2
Link prediction Recommendations
Worked design
Practice
141. Design Personalized News Feed Ranking
Order each home feed so the time is worth it: candidate sources, a multi-task ranker and value model, integrity re-ranking and feedback loops.
ClassicMedium
Machine learning AWS Multi task
+2
Ranking Recommendations
Worked design
Practice
142. Design an Event Recommendation System
Rank upcoming events when every event is new and expires: geo and time candidates, live features, calibrated ranking, a two-sided market.
ClassicMedium
Machine learning AWS Geospatial
+2
Ranking Recommendations
Worked design
Practice
143. Design YouTube Video Search
Find the right videos for a typed query among billions: BM25 and a dual encoder over text, speech and frames, LambdaMART on debiased clicks, human raters.
ClassicMedium
Machine learning Search AWS
+2
Multimodal Ranking
Worked design
Practice
144. Design Personalized Headshot Generation
Selfies in, professional headshots out: per-order LoRA on SDXL, an identity-encoder preview, face-match ranking, GPU pools, consent and deletion.
ClassicMedium
Generative AI AWS Diffusion
+2
Image generation Personalization
Worked design
Practice
145. Design High-Resolution Image Synthesis
Images at 2K–4K without paying for every pixel: latent diffusion, an SR cascade with noise augmentation, tiled decoding, distilled drafts, cost per megapixel: one-hour boards for junior, senior and staff, with the theory.
ClassicMedium
Generative AI AWS Diffusion
+2
Image generation Super resolution
Worked design
Practice
146. Design Text-to-Video Generation
Short clips from a prompt: latent video diffusion, recaptioned data, spacetime attention, cascades, multi-GPU jobs, previews, fair queues, provenance.
ClassicMedium
Generative AI AWS Diffusion
+2
Multimodal Video generation
Worked design
Practice
147. Design Street View Blurring
Blur every face and licence plate in billions of street panoramas: tiles for tiny faces, recall-first detection, a batch pipeline that fails closed.
ClassicMedium
Machine learning Privacy AWS
+2
Computer vision Object detection
Worked design
Practice
148. Design a Text Sharing Service like Pastebin
The URL shortener with a payload: the same id generation and cache-first read path, but the value is kilobytes of text, so it moves out of the database and into object storage.
Product & backendEasy
Caching Object storage Expiry
+1
Pastebin
Worked design
Practice

Each design brings together the brief, practice, examples and discussion.