# Technical glossary Look up a concept, follow its connections, or try it in a local tool. Each definition includes an example, a limitation and a primary source. [Explore the four tools](). ## A2A (Agent2Agent) AI An open protocol through which agents built on different frameworks discover each other's capabilities and exchange tasks and messages. **In practice.** A planning agent reads a data agent's Agent Card and delegates a report as a tracked task. **Watch for.** Interoperability does not establish trust: authentication, authorization and task limits remain the deployer's responsibility. **Related concepts:** [MCP](<#mcp>) · [Multi-agent systems](<#multi-agent>) · [AI agent](<#agent>) [Read the source · A2A Protocol · specification]() [Link to this definition](<#a2a>) ## AI agent AI A system in which a model chooses tools or steps to pursue a task within configured boundaries. **In practice.** Inspect an incident, request a permitted log lookup and propose the next step. **Watch for.** Model output is not authority: software must enforce permissions and approval. **Related concepts:** [MCP](<#mcp>) · [A2A (Agent2Agent)](<#a2a>) · [Durable execution](<#durable-execution>) · [RAG](<#rag>) · [Evaluation](<#evaluation>) [Read the source · Anthropic · building effective agents]() [Link to this definition](<#agent>) [Try it: Replay an agent crash]() ## AIOps Platform Use AI to assist operations, such as alert correlation, incident investigation and remediation proposals. **In practice.** Connect a latency alert with traces and a recent deployment for an operator to review. **Watch for.** Correlation is not proof of cause; remediation needs scoped authority and validation. **Related concepts:** [Site reliability engineering (SRE)](<#sre>) · [Observability](<#observability>) · [DataOps](<#dataops>) · [Evaluation](<#evaluation>) [Read the source · AWS · CloudWatch investigations]() [Link to this definition](<#aiops>) ## Apache Airflow Data Coordinate workflows defined as code, including dependencies, schedules, retries and task state. **In practice.** Run extraction, validation and reporting in dependency order and inspect failed tasks. **Watch for.** It is an orchestrator, not a per-event streaming engine. **Related concepts:** [Batch processing](<#batch>) · [DataOps](<#dataops>) · [dbt](<#dbt>) · [Durable execution](<#durable-execution>) [Read the source · Apache Airflow · documentation]() [Link to this definition](<#airflow>) ## Apache Kafka Data A distributed event-streaming platform with partitioned logs that producers write and consumers can replay. **In practice.** Publish order events so inventory and analytics can consume them independently. **Watch for.** Ordering is per partition, not global; retention and delivery semantics need deliberate configuration. **Related concepts:** [Streaming](<#streaming>) · [Change data capture](<#cdc>) · [Dead-letter queue](<#dlq>) [Read the source · Apache Kafka · introduction]() [Link to this definition](<#kafka>) ## Backfill Data Recomputing or loading data for a past period, for example after fixing a transformation or adding a column. **In practice.** After correcting a currency bug, rerun the daily revenue model for the last 90 partitions. **Watch for.** A backfill must be idempotent and bounded: overwrite whole partitions and limit its load on shared engines. **Related concepts:** [Idempotency](<#idempotency>) · [Batch processing](<#batch>) · [Apache Airflow](<#airflow>) · [dbt](<#dbt>) · [Table snapshot and time travel](<#snapshot>) [Read the source · Apache Airflow · backfill]() [Link to this definition](<#backfill>) ## Backoff Platform Delay successive retries to reduce pressure on a failing dependency; jitter spreads synchronized clients. **In practice.** Try again after progressively longer bounded delays, with a total retry budget. **Watch for.** More retries can amplify an outage. Retry only eligible failures and make side effects safe. **Related concepts:** [Idempotency](<#idempotency>) · [Circuit breaker](<#circuit-breaker>) · [Dead-letter queue](<#dlq>) [Read the source · AWS Builders’ Library]() [Link to this definition](<#backoff>) [Try it: Compare retry policies]() ## Batch processing Data Process a bounded collection of data as a job, rather than treating it as an endless flow. **In practice.** Calculate yesterday's sales from a closed set of orders. **Watch for.** Scheduling frequency and job duration determine how long readers wait for new results. **Related concepts:** [Streaming](<#streaming>) · [Apache Airflow](<#airflow>) · [DuckDB](<#duckdb>) · [Backfill](<#backfill>) [Read the source · Apache Flink · execution modes]() [Link to this definition](<#batch>) ## Change data capture Data Turn database changes into events that downstream consumers can process incrementally. **In practice.** Propagate inserts, updates and deletes from an operational database to an analytical system. **Watch for.** Snapshots, replay, deletes and schema changes require explicit consumer behavior. **Related concepts:** [Streaming](<#streaming>) · [Apache Kafka](<#kafka>) · [Schema evolution](<#schema-evolution>) · [Idempotency](<#idempotency>) [Read the source · Debezium]() [Link to this definition](<#cdc>) ## Circuit breaker Platform Temporarily stop calls after enough failures, then admit limited probes to check recovery. **In practice.** Open a circuit during a persistent dependency outage and test it after a cooldown. **Watch for.** It does not repair the dependency and can delay recovery if thresholds are poorly chosen. **Related concepts:** [Backoff](<#backoff>) · [SLO](<#slo>) · [Site reliability engineering (SRE)](<#sre>) [Read the source · Azure Architecture Center]() [Link to this definition](<#circuit-breaker>) ## ClickHouse Data A column-oriented SQL database designed for analytical queries over large volumes of data. **In practice.** Aggregate event histories to power an operational dashboard. **Watch for.** An analytical serving layer needs its own ingestion, retention and recovery design; it does not replace every transactional database. **Related concepts:** [OLAP](<#olap>) · [Streaming](<#streaming>) · [Apache Kafka](<#kafka>) · [DuckDB](<#duckdb>) [Read the source · ClickHouse · introduction]() [Link to this definition](<#clickhouse>) ## Compaction Data Rewriting many small data files into fewer files near a target size, without changing the table's rows. **In practice.** Five-minute streaming commits leave 11 MiB files; a nightly rewrite merges each hourly partition into one 512 MiB file. **Watch for.** Compaction commits a new snapshot; the small files stay on storage until older snapshots expire. **Related concepts:** [Open table format](<#open-table-format>) · [Table snapshot and time travel](<#snapshot>) · [Streaming](<#streaming>) · [Object storage](<#object-storage>) [Read the source · Apache Iceberg · compact data files]() [Link to this definition](<#compaction>) [Try it: Plan table files and compaction]() ## Data lineage Data Recorded relationships between datasets, jobs and runs that explain where data came from and how it changed. **In practice.** Trace a dashboard field back to the job and source dataset that produced it. **Watch for.** Lineage coverage can be incomplete; a dependency graph does not prove data correctness. **Related concepts:** [DataOps](<#dataops>) · [dbt](<#dbt>) · [Medallion architecture](<#medallion>) [Read the source · OpenLineage]() [Link to this definition](<#lineage>) ## DataOps Data Collaborative practices that improve delivery and operation of analytical data through automation, testing and feedback. **In practice.** Validate a pipeline change with representative data before releasing it and observe quality afterwards. **Watch for.** An orchestrator alone cannot replace ownership, quality criteria or communication with consumers. **Related concepts:** [Data lineage](<#lineage>) · [Apache Airflow](<#airflow>) · [dbt](<#dbt>) [Read the source · IBM · DataOps]() [Link to this definition](<#dataops>) ## dbt Data Build, test and document data transformations from versioned models and their dependencies. **In practice.** Turn raw order tables into tested revenue models with traceable dependencies. **Watch for.** dbt does not replace ingestion; check the v1 or v2 line, the open or product-licensed build and the adapter separately. **Related concepts:** [DataOps](<#dataops>) · [Data lineage](<#lineage>) · [Medallion architecture](<#medallion>) [Read the source · dbt · introduction]() [Link to this definition](<#dbt>) ## Dead-letter queue Platform A separate destination for messages that could not be processed within the configured attempt limit. **In practice.** Quarantine an invalid event after repeated failures and inspect it before replay. **Watch for.** Moving a message to a DLQ is not successful processing; retention and replay need an owner. **Related concepts:** [Backoff](<#backoff>) · [Idempotency](<#idempotency>) · [Apache Kafka](<#kafka>) [Read the source · AWS SQS]() [Link to this definition](<#dlq>) [Try it: Compare retry policies]() ## DevOps Platform Shared practices and responsibility across software delivery and operation, supported by automation and feedback. **In practice.** The team that ships a service also uses incident feedback to improve its delivery process. **Watch for.** Renaming a team or installing a CI server does not establish shared ownership. **Related concepts:** [GitOps](<#gitops>) · [Site reliability engineering (SRE)](<#sre>) · [DevSecOps](<#devsecops>) [Read the source · CNCF Glossary]() [Link to this definition](<#devops>) ## DevSecOps Platform Integrate security practices and shared responsibility throughout software development, delivery and operation. **In practice.** Review threat assumptions, verify dependencies and test authorization before releasing a change. **Watch for.** A scanner passing does not establish secure design; exceptions, remediation and incident learning still need owners. **Related concepts:** [DevOps](<#devops>) · [GitOps](<#gitops>) · [Zero trust](<#zero-trust>) · [MCP](<#mcp>) [Read the source · NIST · SSDF]() [Link to this definition](<#devsecops>) ## DuckDB Data An analytical SQL engine that can run inside an application and query files such as Parquet. **In practice.** Explore a local dataset without deploying a separate database server. **Watch for.** Local performance does not establish distributed or multi-user service capacity; preview features have separate limits. **Related concepts:** [OLAP](<#olap>) · [Batch processing](<#batch>) · [Lakehouse](<#lakehouse>) · [ClickHouse](<#clickhouse>) [Read the source · DuckDB · introduction]() [Link to this definition](<#duckdb>) ## Durable execution Platform Workflow execution whose progress is recorded, so it resumes after crashes, deployments or long waits without repeating completed steps. **In practice.** An approval workflow waits three days for a person and, after a restart, continues at the next step. **Watch for.** Steps with external side effects still need idempotency, and code changes must stay compatible with running histories. **Related concepts:** [Idempotency](<#idempotency>) · [AI agent](<#agent>) · [Apache Airflow](<#airflow>) · [Backoff](<#backoff>) [Read the source · Temporal · understanding Temporal]() [Link to this definition](<#durable-execution>) [Try it: Compare recovery strategies]() ## Embedding AI A fixed-length vector that a model assigns to text, images or other data, so that similar meaning ends up close together. **In practice.** Two support tickets about a failed payment get nearby vectors even when they share no words. **Watch for.** Vectors from different models, versions or dimensions are not comparable; changing the model means re-embedding the corpus. **Related concepts:** [Vector search](<#vector-search>) · [RAG](<#rag>) · [Evaluation](<#evaluation>) [Read the source · Google · Machine Learning Crash Course: embeddings]() [Link to this definition](<#embedding>) ## Error budget Platform The unreliability an SLO allows over its window: one minus the target, applied to eligible requests or measured time. **In practice.** A 99.9% target over 10 million monthly requests allows 10,000 failed requests. **Watch for.** An exhausted budget is a signal to prioritize reliability work, not an automatic penalty; partial windows need care. **Related concepts:** [SLO](<#slo>) · [Site reliability engineering (SRE)](<#sre>) · [Observability](<#observability>) · [Tail latency (p95/p99)](<#tail-latency>) [Read the source · Google SRE · embracing risk]() [Link to this definition](<#error-budget>) ## Evaluation AI A defined set of tasks and scoring rules used to inspect system behavior, including failures. **In practice.** Check task completion and unauthorized actions separately on a held-out test set. **Watch for.** A high aggregate score can hide dangerous failures; model judges also need calibration. **Related concepts:** [RAG](<#rag>) · [AI agent](<#agent>) · [LLMOps](<#llmops>) [Read the source · Anthropic Engineering]() [Link to this definition](<#evaluation>) [Try it: Put an interval around a pass rate]() ## FinOps Platform Shared practices connecting technology spending, ownership and business value through engineering, finance and product collaboration. **In practice.** Attribute a service's cost, review demand and compare cost per successful task. **Watch for.** Reducing a bill without preserving needed quality, reliability or value can be a false saving. **Related concepts:** [Private cloud](<#private-cloud>) · [KV cache](<#kv-cache>) · [SLO](<#slo>) [Read the source · FinOps Foundation · framework]() [Link to this definition](<#finops>) ## GitOps Platform Continuous reconciliation of running systems against a versioned description of their intended state. **In practice.** A controller restores a deployment after an unapproved configuration change. **Watch for.** A pipeline that runs scripts from Git is not necessarily a reconciliation loop. **Related concepts:** [Infrastructure as code](<#iac>) · [DevOps](<#devops>) · [DevSecOps](<#devsecops>) [Read the source · CNCF Glossary]() [Link to this definition](<#gitops>) ## Hybrid attention AI A model design that interleaves full-attention layers with linear-attention layers whose recurrent state does not grow with context. **In practice.** Qwen3.8 keeps a KV cache in one layer of every four; the other three hold a fixed state per request. **Watch for.** The saving applies to the cache, not the weights; long-context recall and runtime support vary by model and engine. **Related concepts:** [KV cache](<#kv-cache>) · [Mixture of experts (MoE)](<#mixture-of-experts>) · [Quantization](<#quantization>) [Read the source · Yang et al., Gated DeltaNet, ICLR 2025]() [Link to this definition](<#hybrid-attention>) ## Idempotency Platform Repeating the same operation has the same intended effect as performing it once. **In practice.** Reuse an operation identifier when retrying an order creation after a timeout. **Watch for.** The key, scope and retention window matter; a timeout does not prove the first attempt failed. **Related concepts:** [Backoff](<#backoff>) · [Dead-letter queue](<#dlq>) · [Durable execution](<#durable-execution>) · [Change data capture](<#cdc>) · [Backfill](<#backfill>) [Read the source · AWS Builders’ Library]() [Link to this definition](<#idempotency>) [Try it: See idempotency keys stop a duplicate write]() ## Infrastructure as code Cloud Infrastructure definitions stored as reviewable, versioned configuration that automation can apply. **In practice.** Review a network change and its plan before applying it. **Watch for.** Version control does not prevent drift or make a destructive plan safe. **Related concepts:** [GitOps](<#gitops>) · [DevOps](<#devops>) · [Private cloud](<#private-cloud>) [Read the source · CNCF Glossary]() [Link to this definition](<#iac>) ## KV cache AI Stored attention keys and values from prior tokens, reused during autoregressive decoding. **In practice.** Reuse the prompt cache while producing each subsequent token. **Watch for.** Memory grows with sequence length and concurrency; hybrid models cache only their full-attention layers, and layout depends on the runtime. **Related concepts:** [Quantization](<#quantization>) · [Hybrid attention](<#hybrid-attention>) · [Prefill and decode](<#prefill-decode>) · [Mixture of experts (MoE)](<#mixture-of-experts>) · [FinOps](<#finops>) [Read the source · Hugging Face Transformers]() [Link to this definition](<#kv-cache>) ## Lakehouse Data An analytical architecture combining lake-style storage with table management and warehouse-style capabilities. **In practice.** Keep analytical tables in object storage and query them through a tested catalog and engine. **Watch for.** Open files alone do not establish transactions, permissions or interoperability. **Related concepts:** [Open table format](<#open-table-format>) · [Object storage](<#object-storage>) · [Query federation](<#query-federation>) · [Medallion architecture](<#medallion>) · [DuckDB](<#duckdb>) · [Data lineage](<#lineage>) [Read the source · Armbrust et al., CIDR 2021]() [Link to this definition](<#lakehouse>) ## Landing zone Cloud A prepared multi-account or multi-project cloud environment with baseline identity, networking, logging and security guardrails for new workloads. **In practice.** A new project inherits central logging, a shared network and organization policies when it is created. **Watch for.** It is a starting baseline, not a finished security posture; its guardrails need owners, exceptions and change control. **Related concepts:** [VPC](<#vpc>) · [Infrastructure as code](<#iac>) · [Zero trust](<#zero-trust>) · [DevSecOps](<#devsecops>) [Read the source · Google Cloud · landing zone design]() [Link to this definition](<#landing-zone>) ## LLMOps AI Lifecycle practices for releasing, evaluating and operating applications built around language models. **In practice.** Version prompts and retrieval data, evaluate a release and observe its latency, cost and failures. **Watch for.** A model benchmark does not validate the application's tools, permissions or real user tasks. **Related concepts:** [MLOps](<#mlops>) · [Evaluation](<#evaluation>) · [KV cache](<#kv-cache>) · [Prefill and decode](<#prefill-decode>) [Read the source · AWS · Generative AI Lens]() [Link to this definition](<#llmops>) ## MCP AI A protocol through which AI applications discover and interact with tools, resources and prompts exposed by servers. **In practice.** Expose a read-only catalog tool through a server with explicit access controls. **Watch for.** Protocol compatibility does not grant permission or make tool output trustworthy. **Related concepts:** [AI agent](<#agent>) · [A2A (Agent2Agent)](<#a2a>) · [Multi-agent systems](<#multi-agent>) · [DevSecOps](<#devsecops>) [Read the source · Model Context Protocol]() [Link to this definition](<#mcp>) ## Medallion architecture Data Organize data by refinement: bronze retains raw input, silver validates it and gold serves a use case. **In practice.** Preserve incoming orders, validate identifiers, then publish a revenue summary. **Watch for.** Layer names are a design convention, not proof of data quality or a requirement to copy every dataset three times. **Related concepts:** [Lakehouse](<#lakehouse>) · [dbt](<#dbt>) · [Data lineage](<#lineage>) [Read the source · Databricks · medallion architecture]() [Link to this definition](<#medallion>) ## Mixture of experts (MoE) AI A model layer that routes each token to a few expert sub-networks, so only part of the parameters computes each token. **In practice.** Qwen3.8-Flash-Next holds 180B parameters but activates about 6B per token. **Watch for.** Every expert stays in memory: the active share lowers compute and bandwidth per token, not resident weight memory. **Related concepts:** [Quantization](<#quantization>) · [KV cache](<#kv-cache>) · [Hybrid attention](<#hybrid-attention>) · [Prefill and decode](<#prefill-decode>) [Read the source · Shazeer et al., ICLR 2017]() [Link to this definition](<#mixture-of-experts>) ## MLOps AI Practices for delivering and operating ML systems across code, data, training, deployment and monitoring. **In practice.** Version a model with its training data references, evaluation results and rollback artifact. **Watch for.** Automating deployment does not ensure model quality or detect every change in production data. **Related concepts:** [LLMOps](<#llmops>) · [Evaluation](<#evaluation>) · [DevOps](<#devops>) [Read the source · Google Cloud Architecture Center]() [Link to this definition](<#mlops>) ## Multi-agent systems AI Several agents cooperate with distinct responsibilities, context or permissions in one workflow. **In practice.** A coordinator requests a report from a data agent and a reviewer checks its sources. **Watch for.** Additional agents increase coordination, latency and failure paths; they do not guarantee better answers. **Related concepts:** [AI agent](<#agent>) · [A2A (Agent2Agent)](<#a2a>) · [MCP](<#mcp>) · [Evaluation](<#evaluation>) [Read the source · Google ADK · workflows]() [Link to this definition](<#multi-agent>) ## Object Lock (WORM) Cloud A storage setting that makes an object version immutable for a retention period: write once, read many. **In practice.** Backups are written with a 30-day lock, so neither ransomware nor an operator mistake can delete them. **Watch for.** A compliance-mode lock cannot be shortened, even by administrators; test retention on a non-production bucket first. **Related concepts:** [Object storage](<#object-storage>) · [Zero trust](<#zero-trust>) · [Private cloud](<#private-cloud>) [Read the source · Amazon S3 · Object Lock]() [Link to this definition](<#object-lock>) ## Object storage Cloud Storage that keeps whole objects with metadata in flat buckets, addressed by key through an HTTP API. **In practice.** Keep Parquet files and table metadata in an S3-compatible bucket that several engines share. **Watch for.** S3 compatibility varies by feature: conditional writes, versioning, object lock and listing behavior must be tested. **Related concepts:** [Open table format](<#open-table-format>) · [Lakehouse](<#lakehouse>) · [Private cloud](<#private-cloud>) · [Object Lock (WORM)](<#object-lock>) [Read the source · Amazon S3 · user guide]() [Link to this definition](<#object-storage>) ## Observability Platform The ability to understand a system from the outside, through the traces, metrics and logs it emits, including problems nobody anticipated. **In practice.** Follow a slow request's trace from the API to the query that caused the delay. **Watch for.** More telemetry is not more understanding; cardinality, sampling and retention decide cost and which questions stay answerable. **Related concepts:** [SLO](<#slo>) · [Site reliability engineering (SRE)](<#sre>) · [AIOps](<#aiops>) · [Error budget](<#error-budget>) · [OpenTelemetry](<#opentelemetry>) [Read the source · OpenTelemetry · observability primer]() [Link to this definition](<#observability>) ## OLAP Data Online analytical processing: queries that scan and aggregate many rows over few columns, usually served by columnar storage and vectorized execution. **In practice.** Aggregate a year of events by region to refresh a dashboard. **Watch for.** Fast scans do not make an engine good at frequent point updates, transactions or high-concurrency writes. **Related concepts:** [ClickHouse](<#clickhouse>) · [DuckDB](<#duckdb>) · [Query federation](<#query-federation>) · [Lakehouse](<#lakehouse>) [Read the source · ClickHouse · what is OLAP?]() [Link to this definition](<#olap>) ## Open table format Data A specification that turns data files in object storage into tables with snapshots, schema and partition metadata that several engines can share. **In practice.** Spark writes an Iceberg table that Trino and DuckDB query through the same catalog. **Watch for.** Engines support different spec versions and features; sharing a format still requires a shared catalog and tested writers. **Related concepts:** [Lakehouse](<#lakehouse>) · [Object storage](<#object-storage>) · [Schema evolution](<#schema-evolution>) · [Query federation](<#query-federation>) · [DuckDB](<#duckdb>) · [Table catalog](<#table-catalog>) [Read the source · Apache Iceberg · table spec]() [Link to this definition](<#open-table-format>) ## OpenTelemetry Platform An open standard, with SDKs and a collector, for producing traces, metrics and logs that share context and naming conventions, exported over OTLP. **In practice.** A request's trace ID follows it from the API through the queue to the worker, so the one slow step shows up end to end. **Watch for.** It standardizes how signals are collected, not where they are stored or alerted on; sampling and label cardinality still decide the cost. **Related concepts:** [Observability](<#observability>) · [Site reliability engineering (SRE)](<#sre>) · [Tail latency (p95/p99)](<#tail-latency>) · [LLMOps](<#llmops>) [Read the source · OpenTelemetry · What is OpenTelemetry?]() [Link to this definition](<#opentelemetry>) ## Prefill and decode AI The two phases of LLM inference: prefill processes the prompt in parallel and is compute-bound; decode generates one token at a time and is memory-bound. **In practice.** Serve long prompts on prefill workers and hand their KV cache to decode workers over a fast interconnect. **Watch for.** Disaggregation adds KV transfer and scheduling cost; it pays off only for some prompt lengths, output lengths and latency targets. **Related concepts:** [KV cache](<#kv-cache>) · [SLO](<#slo>) · [Mixture of experts (MoE)](<#mixture-of-experts>) · [LLMOps](<#llmops>) [Read the source · Zhong et al., DistServe, OSDI 2024]() [Link to this definition](<#prefill-decode>) ## Private cloud Cloud Cloud resources dedicated to one organization, whether operated on its premises or by a provider. **In practice.** An internal platform offers self-service environments to several business units. **Watch for.** Virtual machines alone do not establish a cloud operating model. **Related concepts:** [VPC](<#vpc>) · [Landing zone](<#landing-zone>) · [Infrastructure as code](<#iac>) · [FinOps](<#finops>) [Read the source · NIST SP 800-145]() [Link to this definition](<#private-cloud>) ## Quantization AI Represent selected model values with lower precision to reduce storage and potentially computation. **In practice.** Evaluate 4-bit weights against the original model using a held-out workload. **Watch for.** Smaller weights do not guarantee faster inference or unchanged quality; hardware support matters, and published checkpoints often keep sensitive layers in BF16. **Related concepts:** [KV cache](<#kv-cache>) · [Mixture of experts (MoE)](<#mixture-of-experts>) · [Evaluation](<#evaluation>) · [LLMOps](<#llmops>) [Read the source · Hugging Face Transformers]() [Link to this definition](<#quantization>) ## Query federation Data Querying several data sources from one SQL engine through connectors, without first copying the data into that engine's storage. **In practice.** Trino joins an Iceberg table with a PostgreSQL reference table in one query. **Watch for.** Speed depends on what each connector can push down; federation still loads the sources and does not reconcile their consistency. **Related concepts:** [Open table format](<#open-table-format>) · [OLAP](<#olap>) · [Lakehouse](<#lakehouse>) · [ClickHouse](<#clickhouse>) [Read the source · Trino · connectors]() [Link to this definition](<#query-federation>) ## RAG AI Generation conditioned on information retrieved from an external collection, rather than model parameters alone. **In practice.** Retrieve relevant policy passages and provide them as context for an answer. **Watch for.** Retrieval can miss evidence; supplied context does not guarantee a supported answer. **Related concepts:** [Vector search](<#vector-search>) · [AI agent](<#agent>) · [Evaluation](<#evaluation>) · [LLMOps](<#llmops>) · [Embedding](<#embedding>) [Read the source · Lewis et al., NeurIPS 2020]() [Link to this definition](<#rag>) ## Schema evolution Data Changing a table or message schema while stating which existing readers and writers keep working. **In practice.** Adding an optional column keeps current readers working; dropping a required one breaks them. **Watch for.** Compatibility in one direction does not imply the other; a rename is a removal plus an addition unless the format tracks column IDs. **Related concepts:** [Open table format](<#open-table-format>) · [Change data capture](<#cdc>) · [Data lineage](<#lineage>) · [dbt](<#dbt>) [Read the source · Apache Iceberg · schema evolution]() [Link to this definition](<#schema-evolution>) ## Site reliability engineering (SRE) Platform Apply software engineering to operating services with explicit reliability goals and limited manual toil. **In practice.** Use an error budget to decide whether to ship a change or address reliability first. **Watch for.** Reliability targets reflect user needs and cost; a dashboard alone is not an SRE practice. **Related concepts:** [SLO](<#slo>) · [Error budget](<#error-budget>) · [Observability](<#observability>) · [DevOps](<#devops>) · [AIOps](<#aiops>) [Read the source · Google SRE · introduction]() [Link to this definition](<#sre>) ## SLO Platform A target for a measured service indicator, over a defined population and time window. **In practice.** At least 99.9% of valid requests succeed over a rolling 30-day window. **Watch for.** A target is not a measurement or a contractual guarantee; exclusions must be explicit. **Related concepts:** [Error budget](<#error-budget>) · [Site reliability engineering (SRE)](<#sre>) · [Observability](<#observability>) · [Backoff](<#backoff>) · [Circuit breaker](<#circuit-breaker>) · [Tail latency (p95/p99)](<#tail-latency>) [Read the source · Google SRE]() [Link to this definition](<#slo>) ## Streaming Data Process events as a continuing flow, with explicit rules for time, state and late arrivals. **In practice.** Update available inventory as order events arrive. **Watch for.** Low latency still requires replay, backpressure and duplicate handling; it does not mean instant or lossless. **Related concepts:** [Batch processing](<#batch>) · [Watermark](<#watermark>) · [Apache Kafka](<#kafka>) · [Compaction](<#compaction>) [Read the source · Apache Flink · event time]() [Link to this definition](<#streaming>) ## Table catalog Data The service that maps a table name to its current metadata and makes each commit atomic, so every engine sees the same table state. **In practice.** Spark writes analytics.orders and Trino reads it through the same REST catalog, so the new snapshot is visible to both. **Watch for.** A shared bucket without a shared catalog is not a shared table: engines must speak the same catalog protocol and honor its permissions. **Related concepts:** [Open table format](<#open-table-format>) · [Lakehouse](<#lakehouse>) · [Query federation](<#query-federation>) · [Table snapshot and time travel](<#snapshot>) [Read the source · Apache Iceberg · REST catalog specification]() [Link to this definition](<#table-catalog>) ## Table snapshot and time travel Data An immutable version of a table's state after a commit; a query can read an older snapshot by its ID or timestamp. **In practice.** Compare today's revenue with the table as it was before last night's backfill. **Watch for.** Old snapshots keep their files on storage until they expire, so retention is a cost and compliance decision. **Related concepts:** [Open table format](<#open-table-format>) · [Compaction](<#compaction>) · [Backfill](<#backfill>) · [Table catalog](<#table-catalog>) [Read the source · Apache Iceberg · time travel queries]() [Link to this definition](<#snapshot>) [Try it: Plan the files a table keeps]() ## Tail latency (p95/p99) Platform The response time that only the slowest few percent of requests exceed, such as the 99th percentile. **In practice.** An API averages 80 ms, but its p99 is 1.2 s: one request in a hundred waits more than a second. **Watch for.** Percentiles cannot be averaged across servers or time windows; aggregate the underlying histograms instead. **Related concepts:** [SLO](<#slo>) · [Error budget](<#error-budget>) · [Observability](<#observability>) · [Prefill and decode](<#prefill-decode>) [Read the source · Google SRE · Service level objectives]() [Link to this definition](<#tail-latency>) ## Vector search AI Retrieval that ranks items by the similarity of their embeddings, numeric vectors a model produces from text, images or other data. **In practice.** Find policy passages close in meaning to a question even when they share no keywords. **Watch for.** Approximate indexes trade recall for speed; exact identifiers, filters and freshness often need hybrid keyword search. **Related concepts:** [RAG](<#rag>) · [Evaluation](<#evaluation>) · [AI agent](<#agent>) · [Embedding](<#embedding>) [Read the source · Malkov and Yashunin, HNSW]() [Link to this definition](<#vector-search>) ## VPC Cloud A logically isolated virtual network in a cloud provider, with controlled addressing and connectivity. **In practice.** Separate application subnets from a database subnet and define their routes. **Watch for.** A VPC is a network boundary, not proof of complete workload isolation. **Related concepts:** [Private cloud](<#private-cloud>) · [Zero trust](<#zero-trust>) · [Landing zone](<#landing-zone>) · [Infrastructure as code](<#iac>) · [DevSecOps](<#devsecops>) [Read the source · AWS VPC]() [Link to this definition](<#vpc>) ## Watermark Data A stream-processing signal of event-time progress, used to decide when time-based operations can advance. **In practice.** Close a window after its watermark passes the window end, with a policy for late events. **Watch for.** A watermark is not a guarantee that no older event will ever arrive. **Related concepts:** [Streaming](<#streaming>) · [Apache Kafka](<#kafka>) · [Change data capture](<#cdc>) [Read the source · Apache Flink]() [Link to this definition](<#watermark>) ## Zero trust Cloud A security model that authorizes each request from verified identity, device and context instead of trusting network location. **In practice.** A service reaches a database only with a short-lived workload identity, even inside the same VPC. **Watch for.** Zero trust is an architecture adopted incrementally, not a product; legacy paths often keep implicit trust. **Related concepts:** [VPC](<#vpc>) · [DevSecOps](<#devsecops>) · [Landing zone](<#landing-zone>) · [Private cloud](<#private-cloud>) [Read the source · NIST SP 800-207]() [Link to this definition](<#zero-trust>)