Curated, not collected

Resource library

Books, papers, videos, tools, and repositories that survived contact with the learning process.

Documentation

vLLM

vLLM documentation

Serving architecture, PagedAttention, continuous batching, quantization, speculative decoding, and distributed inference.

AI SystemsInference
IntermediateOpen resource ↗

Paper

USENIX / vLLM

PagedAttention paper

The system design behind efficient KV-cache memory management and high-throughput LLM serving.

AI SystemsInferencePapers
AdvancedOpen resource ↗

Visual guide

Jay Alammar

The Illustrated Transformer

A visual starting point for attention, encoders, decoders, and transformer information flow.

AI SystemsFoundations
FoundationOpen resource ↗

Course

PyTorch

PyTorch tutorials

Official tutorials covering tensors, autograd, neural networks, optimization, distributed training, and performance.

AI SystemsFoundationsTools
FoundationOpen resource ↗

Documentation

Ragas

Ragas documentation

Metrics and workflows for evaluating retrieval-augmented generation systems and test datasets.

AI SystemsEvaluationRAG
IntermediateOpen resource ↗

Documentation

MLflow

MLflow LLM and agent evaluation

Evaluation datasets, scorers, judges, regression testing, tracing, and production monitoring.

AI SystemsEvaluationTools
IntermediateOpen resource ↗

Documentation

Langfuse

Langfuse documentation

Open-source tracing, prompt management, experiments, evaluation, cost, and latency analysis.

AI SystemsObservabilityEvaluationTools
IntermediateOpen resource ↗

Documentation

PostgreSQL

PostgreSQL MVCC

Official explanation of transaction isolation and multi-version concurrency control.

DatabasesPostgres
IntermediateOpen resource ↗

Documentation

PostgreSQL

PostgreSQL index access methods

How PostgreSQL indexes map keys to tuple versions and interact with storage and MVCC.

DatabasesPostgresPerformance
AdvancedOpen resource ↗

Documentation

ClickHouse

ClickHouse documentation

Columnar storage, MergeTree engines, query execution, ingestion, scaling, and operations.

DatabasesClickHouse
IntermediateOpen resource ↗

Documentation

ClickHouse

MergeTree tables

The core ClickHouse table engine: parts, merges, ordering keys, primary indexes, and storage behavior.

DatabasesClickHousePerformance
AdvancedOpen resource ↗

Documentation

Redis

Redis data types

Official guides to strings, hashes, sets, sorted sets, streams, probabilistic types, and their use cases.

DatabasesRedisFoundations
FoundationOpen resource ↗

Architecture guide

Apache Kafka

Apache Kafka design

The commit-log model, batching, partitions, consumers, replication, and delivery guarantees.

Distributed SystemsKafka
AdvancedOpen resource ↗

Book companion

Martin Kleppmann

Designing Data-Intensive Applications resources

Author-maintained references and talks supporting the book's models for data systems and distributed trade-offs.

Distributed SystemsDatabasesFoundations
IntermediateOpen resource ↗

Course

MIT

MIT 6.5840 Distributed Systems

Lectures and labs covering MapReduce, Raft, fault tolerance, replication, and distributed storage.

Distributed SystemsFoundations
AdvancedOpen resource ↗

Documentation

Kubernetes

Kubernetes cluster architecture

Official guide to control-plane components, nodes, workloads, and the reconciliation model.

Distributed SystemsKubernetesFoundations
FoundationOpen resource ↗

Documentation

Kubernetes

Kubernetes observability

Metrics, logs, traces, cluster signals, and end-to-end operational visibility.

Distributed SystemsKubernetesObservability
IntermediateOpen resource ↗

Documentation

OpenTelemetry

OpenTelemetry documentation

Vendor-neutral instrumentation for traces, metrics, logs, context propagation, collectors, and semantic conventions.

Distributed SystemsObservabilityTools
IntermediateOpen resource ↗

Book

Brian Goetz et al.

Java Concurrency in Practice

The foundational model for thread safety, shared state, synchronization, executors, liveness, and performance.

JavaFoundations
IntermediateOpen resource ↗

Article

Aleksey Shipilëv

Java Memory Model pragmatics

Aleksey Shipilëv's practical explanation of ordering, visibility, synchronization, and JMM behavior.

JavaPerformance
AdvancedOpen resource ↗

Showing 20 of 20 curated learning resources