Create a Config

Build a reconify.yaml that tells Reconify where your files are and how to parse them.

Every Reconify run reads a single reconify.yaml file. That file defines your sources (where files live and how to parse them), your pairs (which two sources to compare and under what rules), and optional index settings for large files. Getting the config right once means every subsequent run is a single command.

What you'll learn

By the end of this guide, you know:

the minimum fields required to describe a source and a pair,
how parsers map source-specific column names to Reconify's normalized transaction model,
and how to validate your config before running a reconciliation.

How it works

Reconify's config is declarative: you describe your data, and Reconify handles parsing, normalization, and matching. Each source maps to one parser block, and each pair links two sources. See Overview for how those pieces fit into the full pipeline.

Command

Prefer a faster start?

Run reconify config init to generate this file interactively from your sample data instead of writing it by hand. See reconify config init.

reconify config validate --config reconify.yaml

Steps

Create `reconify.yaml`

Start with the minimum shape: two sources and one pair.

version: 1
timezone: "UTC"

sources:
  bank:
    file_pattern: "data/bank.csv"
    parser:
      type: csv
      date_col: "Date"
      date_layout: "2006-01-02"
      amount_col: "Amount"
      decimal: "."
      thousands: ","
      multiplier: 100
      ref_col: "Reference"
      name_col: "Description"

  ledger:
    file_pattern: "data/ledger.csv"
    parser:
      type: csv
      date_col: "posted_at"
      date_layout: "2006-01-02"
      amount_col: "amount_minor"
      multiplier: 1
      ref_col: "payment_id"
      name_col: "memo"

pairs:
  bank_vs_ledger:
    left: bank
    right: ledger
    date_window: "1d"
    amount_tolerance_minor: 0
    name_mode: "none"

multiplier converts source amounts to minor units. Use 100 when the file contains 123.45 (major units). Use 1 when it already contains 12345 (minor units, as in Stripe's API output).

Add index settings for large files

If your right-side file can exceed a few hundred MB, add an index block so Reconify spills to disk automatically instead of running out of memory.

index:
  backend: auto
  spill_dir: "/tmp/reconify"
  auto_max_right_file_mb: 2048

Validate the config

Validation catches missing required fields, unknown sources in pairs, invalid parser types, invalid timezones, invalid date window units, and negative tolerances.

reconify config validate --config reconify.yaml

Fix any reported errors before proceeding.

Verify it worked

A passing validate run prints:

reconify.yaml is valid

If anything is wrong, the error message names the specific field. The most common issues are listed below.

multiplier is 0, which is not valid. It must be a positive integer.
date_layout uses the wrong reference date. Reconify uses Go's reference time (2006-01-02, not YYYY-MM-DD).
A pair references a source name that does not exist in sources.

Once the config is valid, run reconify config check-source to confirm your actual input files match the parser settings.

On this page