Free SDK

Kanoniv's Python SDK and embedded Rust reconciliation engine are free to use locally. No account, no API keys, no data leaves your machine. You can validate specs, plan execution, diff changes, and reconcile identity data entirely offline.

What's Free (Local SDK)

Component	Description
Python SDK (pip install kanoniv)	Validate, plan, diff, and reconcile locally
Rust reconciliation engine	Native-speed matching, blocking, scoring, and survivorship
All matching capabilities	Exact, similarity, range, composite rules with no restrictions
All blocking strategies	Plan-based and LSH blocking
All survivorship policies	most_recent, most_complete, source_priority, aggregate
CLI offline commands	kanoniv validate, kanoniv plan, kanoniv diff

What's Cloud (Paid)

Component	Description
Kanoniv Cloud	Managed platform with persistent identity graph
Auto-discovery + bootstrap	kanoniv autodetect --bootstrap via Cloud API
AutoTune	Automated threshold optimization via Cloud API
Real-time resolution	Sub-millisecond resolve API
Warehouse connectors	Snowflake, Databricks, BigQuery
Enterprise features	SSO, SCIM, HIPAA compliance, BYOK encryption, SIEM, audit logs

Architecture

The SDK has zero external dependencies at runtime - no API calls, no telemetry, no data leaves your machine.

Your Python code
     |
kanoniv Python SDK (pip install kanoniv)
     |
PyO3 bridge (Python -> Rust FFI)
     |
Rust reconciliation engine (compiled native code)
     |
Results (golden records, match pairs, diagnostics)

The Rust engine is compiled into a native Python extension via PyO3 and maturin. When you pip install kanoniv, you get a pre-built wheel for your platform - no Rust toolchain needed.

Supported Platforms

Platform	Architecture	Wheel
Linux	x86_64	kanoniv-*-manylinux_2_17_x86_64.whl
macOS	Intel (x86_64)	kanoniv-*-macosx_10_12_x86_64.whl
macOS	Apple Silicon (aarch64)	kanoniv-*-macosx_11_0_arm64.whl

Installation

pip install kanoniv

Or with Cloud support:

pip install kanoniv[cloud]

Quick Example

# customer-spec.yaml
entity:
  name: customer
sources:
  - name: crm
    adapter: csv
    location: contacts.csv
    primary_key: id
  - name: billing
    adapter: csv
    location: invoices.csv
    primary_key: id
rules:
  - name: email_exact
    type: exact
    field: email
    weight: 1.0
  - name: name_fuzzy
    type: jaro_winkler
    field: name
    threshold: 0.9
    weight: 0.8
survivorship:
  strategy: source_priority
  priority: [crm, billing]
decision:
  thresholds:
    match: 0.85

from kanoniv import Spec, Source, reconcile, validate

# Validate the spec (catches errors before processing)
spec = Spec.from_file("customer-spec.yaml")
validation = validate(spec)
validation.raise_on_error()

# Load sources
sources = [
    Source.from_csv("crm", "contacts.csv"),
    Source.from_csv("billing", "invoices.csv"),
]

# Run reconciliation (entirely local - no network calls)
result = reconcile(sources, spec)

print(f"Input records: {result.total_input_records}")
print(f"Golden records: {len(result.golden_records)}")
print(f"Merge rate: {result.merge_rate:.1%}")

Source Adapters

The SDK includes adapters for common data sources:

Adapter	Method	Description
CSV	Source.from_csv(name, path)	Read from CSV files
Pandas	Source.from_pandas(name, df)	Read from DataFrames
Warehouse	Source.from_warehouse(name, conn, query)	Read from SQL databases
dbt	Source.from_dbt(name, project, model)	Read from dbt models

Why Free?

For users

• No vendor lock-in: Your matching logic lives in a YAML file you own. If you stop using Kanoniv, your spec is still valid documentation.
• Transparency: The matching engine produces explainable decisions with confidence scores and rule traces.
• Offline-first: Develop and test locally with real data before deploying to production. No account, no API key, no internet required.
• Free for any scale: Process millions of records locally at zero cost. The Cloud is optional.

For the ecosystem

• Interoperability: The YAML spec format is a portable standard for defining matching rules. Other tools can read and write it.
• Trust: Every merge decision has a reason code, confidence score, and field provenance - critical for compliance and regulated industries.