← projects

[Ledger Glue]

Finance Integrations

Corporate-card transactions into accounting, a multi-provider bank ETL with central-bank FX, and projection spreadsheets generated by code.

● ShippedData EngineeringClient workPythonAWS LambdaPostgresGoogle Sheets API
01 /context

The finance team reconciled cards, bank accounts and projections by hand across tools that didn't talk to each other.

02 /my role

Solo: I designed and built the three integrations for the finance team, from the API clients to deployment and scheduling.

03 /problem

Copy-paste between card portals, banks and the accounting system, with FX conversion done by hand.

04 /how it works
  1. 01 · Card → accounting

    A serverless function runs every hour, fetches corporate-card transactions and posts them to the accounting software per card; credentials come from a parameter store.

  2. 02 · Bank ETL

    One client per provider (two digital banks, a payments platform, a corporate card) behind a common interface, plus central-bank FX rates, loaded into Postgres.

  3. 03 · Projections as code

    OAuth scripts create and format a long-horizon projection model (an 18-year forestry cycle) directly in Google Sheets — formulas, currency formats, frozen panes.

05 /architecture
01
Card + bank providers
02
Serverless sync (hourly)
03
Central-bank FX
04
Postgres
05
Accounting · spreadsheets
06 /results
Card transactions reach accounting within the hour, with no manual import.
07 /related