Apache APISIX Gateway

Spotify's Annoy library indexes high-dimensional vectors in memory. Fast search and low memory usage. Python and C++ implementations. Used internally by Spotify. Active maintenance.

Apache Cassandra stores time-series at petabyte scale. Write-heavy workload optimized. Time-bucketing for efficient queries. Replication across regions. Used by Apple and Netflix.

Chroma is an open-source embedding database built for AI applications. Run locally or distributed. SQLite backend. Hugging Face integration. Simple API. Easy to get started.

eBPF programs run safely in the Linux kernel. Monitor system calls, network, disk. No recompile needed. Used by Cilium, Falco, and Pixie for observability. New programming model transforming Linux inf…

Meta's Faiss library searches billions of vectors. GPU acceleration with CUDA. Index compression reduces memory. Research and production use. Widely used in recommendation systems.

Freebase is a large collaborative knowledge base. Now part of Google Knowledge Graph. CC licensed, freely available for download. Groundtruth for knowledge graph research. 43M+ entities extracted.

GQL is a standardized graph query language (ISO/IEC 39075). Successor to Cypher. Multi-vendor support planned. Early adoption by industry players. Will unify graph query ecosystem.

Grafana Loki is a horizontally scalable log aggregation system. Label-based indexing stores logs cost-effectively. LogQL queries filter by service, pod, region. No high cardinality concerns. Pairs wit…

KubeVirt lets you run virtual machines on Kubernetes like pods. Useful for legacy VMs or Windows workloads. Networking and storage APIs consistent. Live migration. Operated as a DaemonSet. Community p…

Marqo is an open-source tensor search engine. No API calls to embeddings service. Local document indexing. Query-specific fine-tuning. Built for ease of use.

Milvus is an open-source vector database for large-scale similarity search. Billion-vector scale. Multiple index types: IVF, HNSW, DiskANN. Cloud-hosted or self-hosted. Supports multiple languages. CN…

Desktop app for running and chatting with local AI models with RAG, web search, and model management.

NMSLIB provides approximate nearest neighbor search. C++, Python, Java, Ruby bindings. HNSW and other algorithms. High performance tuning options. Research origins.

OpenTelemetry is a vendor-neutral standard for collecting metrics, traces, and logs from any application. Collector receives data from SDKs, transforms, and exports to backends (Datadog, Grafana, Splu…

OpenTSDB stores time-series on top of HBase. Billions of metrics at millisecond precision. Tag-based queries. Built-in aggregators for rollups. Java-based backend.

One-click installer for AI applications that sets up Stable Diffusion, LLMs, and other AI tools locally.

Ponchao is Tencent's open-source chaos testing framework. Multi-platform support (cloud, on-premise). Orchestrates complex scenarios. Real-time status monitoring. Growing adoption in Asia.

pgvector is an open-source extension for Postgres. Store and search vectors in Postgres. Index types: IVF, HNSW. No separate database needed. Simple to deploy. Community-maintained.

Prometheus scrapes metrics from HTTP endpoints every 15 seconds. Time-series with labels enable multi-dimensional queries. Pull-based avoids overwhelming servers. AlertManager routes incidents. De-fac…

Qdrant is an open-source vector database optimized for semantic search and recommendation systems. HNSW indexing with pruning. Payload storage with filtering. Snapshots and recovery. Rust implementati…

StatsD is a lightweight protocol and reference implementation for publishing application metrics. Applications send counters, timers, and gauge values via UDP packets to a local agent. The agent aggre…

Telegraf is a plugin-driven server agent for collecting metrics. 200+ input plugins (CPU, disk, Docker, Prometheus). Output to InfluxDB, Graphite, or Kafka. Lightweight, single binary. Standard in mon…

Thanos is a set of components extending Prometheus. Sidecar uploads blocks to S3. Querier aggregates across all Prometheus instances. 5-year retention. Ruler for alert generation. CNCF project.

Vald is an open-source distributed vector database. High-dimensional approximate nearest neighbor search. Horizontally scalable. Python and Go clients. Japanese origin, growing adoption.

Velero backs up Kubernetes resources and persistent volumes to cloud storage (S3, GCS, Azure). Disaster recovery: restore to new cluster in minutes. Migration tool for multi-cluster ops. Hooks for dat…