# Golden Suite

> A polyglot data-quality and entity-resolution toolkit. Zero-config, AI-native, MIT-licensed.

## Docs

- [Zero-config vs. hand-tuned Splink](https://docs.bensevern.dev/docs/case-studies/zero-config-vs-splink.md): A fair, deterministic bake-off: GoldenMatch with zero tuning beats an expert hand-rolled Splink model on pairwise F1 across every PII dataset Splink scores — judged by one shared evaluator.
- [Architecture](https://docs.bensevern.dev/docs/concepts/architecture.md): How the six Golden Suite tools compose into one pipeline across Python, TypeScript, Rust, dbt, and GitHub Actions.
- [Cross-language parity & phase-handoff limits](https://docs.bensevern.dev/docs/concepts/cross-language-parity.md): What actually round-trips between the Python and TypeScript ports of GoldenMatch — which pipeline phases you can hand off byte-for-byte, which are tolerance-bounded, and which can't cross at all.
- [Entity resolution](https://docs.bensevern.dev/docs/concepts/entity-resolution.md): The core concepts behind deduplication and record linkage: blocking, scoring, clustering, and survivorship.
- [Scale envelope](https://docs.bensevern.dev/docs/concepts/scale-envelope.md): Pick the right backend for your row count, and avoid the block-size failure modes that dominate ER performance.
- [SQL extensions](https://docs.bensevern.dev/docs/extensions/sql.md): Run GoldenMatch fuzzy matching directly inside PostgreSQL and DuckDB.
- [For researchers](https://docs.bensevern.dev/docs/for-researchers.md): Reproduce the benchmarks, read the evaluation methodology and honest framing, cite the project, and find the comparison against other entity-resolution tools.
- [Analyzers](https://docs.bensevern.dev/docs/goldenanalysis/analyzers.md): The GoldenAnalysis analyzer catalog: the read-only metrics each one computes, how a run is configured, and the metric + regression-gating model.
- [GoldenAnalysis CLI](https://docs.bensevern.dev/docs/goldenanalysis/cli.md): Every GoldenAnalysis command for reporting, trending, and regression detection.
- [Config matrix](https://docs.bensevern.dev/docs/goldenanalysis/config-matrix.md): The full matrix of GoldenAnalysis config options and vocabularies -- generated so it never drifts.
- [Cross-run analysis](https://docs.bensevern.dev/docs/goldenanalysis/cross-run.md): Trend metrics across a run history and flag regressions versus a baseline, with a deterministic narrative.
- [Native accelerator](https://docs.bensevern.dev/docs/goldenanalysis/native.md): An optional Rust kernel for the heavy aggregation primitives, gated only after a measured wall-clock win.
- [GoldenAnalysis overview](https://docs.bensevern.dev/docs/goldenanalysis/overview.md): Read-only, cross-cutting analysis, metrics, and reporting across the Golden Suite. Consumes any stage's artifacts and emits one unified report.
- [Analysis recipes](https://docs.bensevern.dev/docs/goldenanalysis/recipes.md): Copy-paste GoldenAnalysis snippets for common jobs: score a dedupe run, roll a whole pipeline into one report, and gate CI on a cross-run regression.
- [Checks](https://docs.bensevern.dev/docs/goldencheck/checks.md): The full catalog of GoldenCheck check types: schema, nullability, keys, ranges, patterns, temporal, relations, referential integrity, denial constraints, and drift.
- [GoldenCheck CLI](https://docs.bensevern.dev/docs/goldencheck/cli.md): Commands, key flags, domain packs, and the goldencheck.yml config format.
- [Config matrix](https://docs.bensevern.dev/docs/goldencheck/config-matrix.md): The full matrix of GoldenCheck config knobs -- generated from the schema so it never drifts.
- [GoldenCheck integrations](https://docs.bensevern.dev/docs/goldencheck/integrations.md): Run GoldenCheck as dbt tests and as a GitHub Action that gates pull requests on data-quality regressions.
- [Native acceleration & deep profiling](https://docs.bensevern.dev/docs/goldencheck/native.md): GoldenCheck's optional Rust/Arrow runtime and the deep-profiling checks it powers: composite keys, functional dependencies, fuzzy values, approximate-FD violations, plus --deep, referential integrity, and freshness.
- [GoldenCheck overview](https://docs.bensevern.dev/docs/goldencheck/overview.md): Zero-config data-quality scanning that discovers rules from your data instead of making you write them.
- [Validation recipes](https://docs.bensevern.dev/docs/goldencheck/recipes.md): Copy-paste GoldenCheck configs for common jobs: validate a customer file, enforce enums and ranges, and catch outliers.
- [GoldenFlow CLI](https://docs.bensevern.dev/docs/goldenflow/cli.md): Commands, flags, and the goldenflow.yaml config format.
- [Config matrix](https://docs.bensevern.dev/docs/goldenflow/config-matrix.md): The full matrix of GoldenFlow config knobs and its transform vocabulary -- generated so it never drifts.
- [Native acceleration](https://docs.bensevern.dev/docs/goldenflow/native.md): GoldenFlow's optional compiled Rust/Arrow runtime — which components run native, the `GOLDENFLOW_NATIVE` gate, and how parity is enforced. Generated from the native loader; do not edit by hand.
- [GoldenFlow overview](https://docs.bensevern.dev/docs/goldenflow/overview.md): Standardize, reshape, and normalize messy data with 92 built-in transforms across 11 categories.
- [Performance](https://docs.bensevern.dev/docs/goldenflow/performance.md): How GoldenFlow stays fast: vectorized Polars fast paths with a per-row fallback, and an optional Arrow-native Rust kernel for phone normalization.
- [Transform recipes](https://docs.bensevern.dev/docs/goldenflow/recipes.md): Copy-paste GoldenFlow configs for common jobs: standardize contact data, clean and dedupe, and reshape columns.
- [Transforms](https://docs.bensevern.dev/docs/goldenflow/transforms.md): The full GoldenFlow transform-op catalog: 124 built-in ops for case, encoding, names, phone, email, dates, addresses, identifiers, phonetics, and more, plus domain packs and scalar canonicalizers.
- [ER agent](https://docs.bensevern.dev/docs/goldenmatch/agent.md): GoldenMatch as an autonomous entity resolution agent: A2A protocol, 47 skills, confidence-gated review queue, and Python API.
- [GoldenMatch API quick reference](https://docs.bensevern.dev/docs/goldenmatch/api-quick-reference.md): Practical examples for the most-used surface. Authoritative type signatures live in goldenmatch/_api.py and goldenmatch/config/schemas.py.
- [Auto-config](https://docs.bensevern.dev/docs/goldenmatch/auto-config.md): How the introspective AutoConfig controller detects column types, picks scorers and blocking, and iterates until it converges.
- [Backends and scale](https://docs.bensevern.dev/docs/goldenmatch/backends-and-scale.md): The polars-direct, bucket, chunked, DuckDB, and Ray backends, and the measured scale numbers for each.
- [Blocking strategies](https://docs.bensevern.dev/docs/goldenmatch/blocking.md): Reduce the O(N²) comparison space with GoldenMatch's 10 blocking strategies: static, adaptive, sorted neighborhood, multi-pass, ANN hybrid, ANN, canopy, learned, MinHash/LSH, and SimHash.
- [GoldenMatch CLI](https://docs.bensevern.dev/docs/goldenmatch/cli.md): Every GoldenMatch command and the key flags for the dedupe pipeline.
- [Config linter](https://docs.bensevern.dev/docs/goldenmatch/config-linter.md): Pre-flight checks the config linter runs against your data shape before a dedupe/match run — what each rule flags, when it fires, and why. Generated from the rule registry; do not edit by hand.
- [Config matrix](https://docs.bensevern.dev/docs/goldenmatch/config-matrix.md): The full matrix of GoldenMatch config knobs, their valid values, and how they combine -- generated from the schema so it never drifts.
- [Config suggestions (the healing loop)](https://docs.bensevern.dev/docs/goldenmatch/config-suggestions.md): GoldenMatch's iterative workflow: zero-config gets good results and returns the config it chose; the healer reviews the results and suggests self-verified tweaks; you apply them, results improve, repeat.
- [Configuration](https://docs.bensevern.dev/docs/goldenmatch/configuration.md): Full YAML reference for GoldenMatch: matchkeys, blocking, golden rules, standardization, validation, LLM scorer, Learning Memory, output, and backends.
- [Data-quality–aware matching](https://docs.bensevern.dev/docs/goldenmatch/data-quality.md): How GoldenMatch uses GoldenCheck's data-quality signal to improve entity-resolution results, recall, precision, and trust — via fail-open, opt-in bridges.
- [Document ingest](https://docs.bensevern.dev/docs/goldenmatch/documents.md): Run GoldenMatch on unstructured input — extract matchable records from PDFs and images, then dedupe them.
- [Domain packs](https://docs.bensevern.dev/docs/goldenmatch/domain-packs.md): Built-in and custom YAML rulebooks that extract structured fields from unstructured product descriptions and other domain-specific text.
- [Evaluation](https://docs.bensevern.dev/docs/goldenmatch/evaluation.md): Measure matching accuracy against ground truth and enforce quality gates in CI/CD pipelines.
- [Identity graph](https://docs.bensevern.dev/docs/goldenmatch/identity-graph.md): Durable, queryable identity graph with stable entity IDs, evidence edges, and a consistent JSON view across Python, SQL, REST, MCP, A2A, and the web UI.
- [Installation](https://docs.bensevern.dev/docs/goldenmatch/installation.md): Install GoldenMatch from PyPI, Docker, or as a PostgreSQL/DuckDB extension. Optional extras add embeddings, LLM scoring, database sync, and more.
- [Knowledge-graph framework integrations](https://docs.bensevern.dev/docs/goldenmatch/kg-integrations.md): Drop goldenmatch in as the entity-resolution stage of neo4j-graphrag, LlamaIndex PropertyGraphIndex, and Graphiti via the goldenmatch-kg package.
- [Learning memory](https://docs.bensevern.dev/docs/goldenmatch/learning-memory.md): Persist steward corrections, unmerge decisions, and LLM votes across runs so the same false positive never comes back.
- [LLM integration](https://docs.bensevern.dev/docs/goldenmatch/llm.md): Use GPT-4o-mini or Claude to score borderline pairs that fuzzy matching alone cannot resolve, with budget caps, model tiering, and iterative calibration.
- [MCP server](https://docs.bensevern.dev/docs/goldenmatch/mcp.md): Connect GoldenMatch to Claude Desktop, Claude Code, and other MCP clients — 85 tools for autonomous entity resolution, data inspection, Learning Memory, and Identity Graph.
- [Migrating from hand-rolled dbt](https://docs.bensevern.dev/docs/goldenmatch/migrating-from-dbt.md): Distill a hand-rolled dbt entity-resolution pipeline down to one reusable GoldenMatch config — and prove it reproduces your existing clusters.
- [Migrating from Splink](https://docs.bensevern.dev/docs/goldenmatch/migrating-from-splink.md): Convert a Splink model to a GoldenMatch config, verify it reproduces Splink's clustering, and run it — in one command or one line of Python.
- [Migrating to v2.0](https://docs.bensevern.dev/docs/goldenmatch/migrating-to-v2.md): Step-by-step guide for upgrading from GoldenMatch 1.x to 2.0.
- [Migrating to v3](https://docs.bensevern.dev/docs/goldenmatch/migrating-to-v3.md): GoldenMatch 3.0.0: Arrow-native results and the Arrow frame backend by default
- [Native acceleration](https://docs.bensevern.dev/docs/goldenmatch/native.md): GoldenMatch's optional compiled Rust/Arrow runtime — which components run native, the `GOLDENMATCH_NATIVE` gate, and how parity is enforced. Generated from the native loader; do not edit by hand.
- [GoldenMatch overview](https://docs.bensevern.dev/docs/goldenmatch/overview.md): Zero-config entity resolution for Python and TypeScript: fuzzy, exact, probabilistic, and LLM scoring with golden-record synthesis.
- [Pipeline](https://docs.bensevern.dev/docs/goldenmatch/pipeline.md): GoldenMatch's 10-step pipeline from raw files to golden records, with per-step API references.
- [Privacy-preserving linkage](https://docs.bensevern.dev/docs/goldenmatch/pprl.md): Match records across organizations without sharing raw data, using Bloom-filter PPRL.
- [Python API](https://docs.bensevern.dev/docs/goldenmatch/python-api.md): Complete reference for all 194 symbols exported by GoldenMatch from a single import.
- [GoldenMatch quickstart](https://docs.bensevern.dev/docs/goldenmatch/quickstart.md): Dedupe a file, match two files, write golden records, and configure matchkeys in Python and TypeScript.
- [Configuration recipes](https://docs.bensevern.dev/docs/goldenmatch/recipes.md): Copy-paste GoldenMatch configs for common jobs: customer dedupe, cross-source linking, typo-heavy names, addresses, image dedupe, 100M distributed, and survivorship.
- [Reference data](https://docs.bensevern.dev/docs/goldenmatch/reference-data.md): The five bundled reference-data packs that auto-config picks up automatically: surnames, given names, business, addresses, and NAICS industry codes.
- [REST API](https://docs.bensevern.dev/docs/goldenmatch/rest-api.md): Local HTTP server for real-time matching, cluster browsing, and data steward review.
- [Scoring](https://docs.bensevern.dev/docs/goldenmatch/scoring.md): All GoldenMatch scoring methods: exact, fuzzy, probabilistic, LLM, embedding, and parallel scoring.
- [Semantic-layer key integrity](https://docs.bensevern.dev/docs/goldenmatch/semantic-key-integrity.md): Certify the entity keys a semantic model (dbt/MetricFlow, Cube, OSI) declares — quantify metric fan-out and entity undercount before you trust a SUM or COUNT(DISTINCT). Advisory, never mutating.
- [Streaming & incremental](https://docs.bensevern.dev/docs/goldenmatch/streaming.md): Single-record matching, micro-batch streaming, and CLI-based incremental matching against existing data in real time.
- [Interactive TUI](https://docs.bensevern.dev/docs/goldenmatch/tui.md): Gold-themed terminal UI for GoldenMatch: 8 tabs for data profiling, configuration, match review, golden records, active learning, export, AutoConfigController telemetry, and Learning Memory corrections.
- [Tuning & opt-ins](https://docs.bensevern.dev/docs/goldenmatch/tuning.md): The single reference for every GoldenMatch runtime knob: native acceleration, backend selection, the distributed pipeline, perf opt-ins, and every GOLDENMATCH_* environment variable — what each does, its default, and when to use or avoid it.
- [TypeScript API](https://docs.bensevern.dev/docs/goldenmatch/typescript.md): GoldenMatch as an npm package with full feature parity with the Python toolkit: edge-safe core, Node-only additions, and all scoring, blocking, and golden-record strategies.
- [What changed from 1.0 to 2.0](https://docs.bensevern.dev/docs/goldenmatch/v1-to-v2.md): The full arc from GoldenMatch 1.0 (March 2026) to 2.0 (June 2026): the capabilities that landed across the 1.x line, plus the four breaking removals in the 2.0 major.
- [v1 vs v2 at a glance](https://docs.bensevern.dev/docs/goldenmatch/v1-vs-v2.md): One-screen comparison of GoldenMatch 1.0 and 2.0: what breaks in the 2.0 major, and the capability deltas that accumulated across the 1.x line.
- [GoldenPipe CLI](https://docs.bensevern.dev/docs/goldenpipe/cli.md): Every GoldenPipe command for running and inspecting pipelines.
- [Config matrix](https://docs.bensevern.dev/docs/goldenpipe/config-matrix.md): The full matrix of GoldenPipe pipeline-config knobs and vocabularies -- generated so it never drifts.
- [GoldenPipe overview](https://docs.bensevern.dev/docs/goldenpipe/overview.md): The orchestrator that chains data-quality checking, transformation, and deduplication into a single adaptive pipeline.
- [Pipeline recipes](https://docs.bensevern.dev/docs/goldenpipe/recipes.md): Copy-paste GoldenPipe configs that wire the suite together: full clean-and-dedupe, profile-and-report, and dedupe-then-resolve-identities.
- [Stages](https://docs.bensevern.dev/docs/goldenpipe/stages.md): The GoldenPipe stage and adapter catalog: every built-in stage, the PipelineConfig / StageSpec shape, status enums, and how a pipeline is wired end to end.
- [Golden Suite](https://docs.bensevern.dev/docs/index.md): A polyglot data-quality and entity-resolution toolkit. Zero-config, AI-native, MIT-licensed.
- [InferMap CLI](https://docs.bensevern.dev/docs/infermap/cli.md): Every InferMap command for schema mapping and inspection.
- [Config matrix](https://docs.bensevern.dev/docs/infermap/config-matrix.md): The full matrix of InferMap config options and vocabularies -- generated so it never drifts.
- [Mapping](https://docs.bensevern.dev/docs/infermap/mapping.md): How InferMap scores and assigns source columns to a target schema: the scorer pipeline, weighted combination, MapEngine configuration, and the inference vocabularies.
- [Native acceleration](https://docs.bensevern.dev/docs/infermap/native.md): InferMap's optional compiled Rust/Arrow runtime — which components run native, the `INFERMAP_NATIVE` gate, and how parity is enforced. Generated from the native loader; do not edit by hand.
- [InferMap overview](https://docs.bensevern.dev/docs/infermap/overview.md): An inference-driven schema mapping engine that aligns messy source columns to a target schema with confidence scores and reasoning.
- [Mapping recipes](https://docs.bensevern.dev/docs/infermap/recipes.md): Copy-paste InferMap snippets for common jobs: map a messy export to your canonical schema, reuse a saved mapping as a validation gate, and add a custom alias or domain pack.
- [Quickstart](https://docs.bensevern.dev/docs/quickstart.md): Deduplicate a CSV in 30 seconds, then run the full pipeline.
- [API surface](https://docs.bensevern.dev/docs/reference/api-surface.md): One page for the whole suite's programmatic surface — Python, TypeScript, CLI, MCP, REST, and agent skills across all six packages, with the primary entry points for each.
- [ER vendor comparison](https://docs.bensevern.dev/docs/reference/vendor-comparison.md): How GoldenMatch compares to other entity-resolution engines across OSS, identity-graph, cloud-managed, and enterprise MDM tiers.
- [Versioning policy](https://docs.bensevern.dev/docs/reference/versioning.md): How the suite versions its Python and TypeScript packages — independent semver per surface, why there is no lockstep, and how cross-surface parity is actually guaranteed.
- [Amortized Bayesian ER (exploratory)](https://docs.bensevern.dev/docs/research/amortized-bayesian-er.md): A documented research arc exploring a novel entity-resolution design: an amortized, calibrated neural posterior over the partition, with a learned reconstruction likelihood and EIG-driven active labelling. Honest outcome: mechanisms validated, not accuracy-competitive.
- [Landscape-sculpting ER (exploratory)](https://docs.bensevern.dev/docs/research/landscape-sculpting-er.md): A second exploratory arc reframing entity resolution as sculpting a potential/attractor landscape. Honest outcome: the framing is genuinely novel but the mechanism is cosmetic — it gives the same partition as a discrete split/merge loop once the objective is calibrated.
- [Suite surface matrix](https://docs.bensevern.dev/docs/suite-matrix.md): Cross-package view of the Golden Suite: capability counts, the Rust -core compute substrate, Python/TypeScript parity, and auto-detected gaps — generated and gated in CI.
- [Repo thesis weaknesses](https://docs.bensevern.dev/docs/thesis-weaknesses.md): Where the codebase does not yet meet the governing architecture frame (one product, two engines, many surfaces) — the thesis-conformance weakness board and live-harvested drift signals, generated and gated in CI.
- [Web workbench (no code)](https://docs.bensevern.dev/docs/web-ui.md): The point-and-click way to use GoldenMatch: edit rules, preview matches, label pairs, and compare runs in a browser — no code after the first install.