Documentation

This Knowledge Base documents TxFlow: ordering it, installing it on Google Cloud or on your own infrastructure, calling its API, and how its components fit together.

TxAgent is not documented here. For what it does, see the TxAgent product page; for anything beyond it, get in touch.

Getting started

In the order a new deployment needs them, rather than alphabetically.

• Order it — request a licence, and get the application from Google Kubernetes Engine.
• Install it (GCP) — deploy TxFlow as a Kubernetes application, then reach the API, including forwarding its port to a local environment.
• Install it (non GCP) — on-premise and other clouds: prerequisites, the Helm chart, and running it as a standalone application.
• Use it — the functional and technical API endpoints, the JSON request shape, worked examples with test data, and how licence and billing enforcement behaves.
• Maintain it — updating images, scaling, and removing the deployment on GCP or elsewhere. TxFlow is a stateless API, so there is no database to back up.
• Pricing and support — the 30-day free trial, how it can be extended, and what follows a proof of concept.

Architecture

• Components — the four functional parts, message parser, enricher, address parser and geo locator, and the API endpoint in front of them.
• Cloud integration — TxFlow ships as a container, so the same artifact runs on GCP, on-premise or on another cloud. What changes is the surrounding plumbing.

Admin guide

• Administrator’s guide — runtime configuration: the standard keys you will set, and the advanced ones you probably should not.

Background

Context rather than instructions — useful before a deployment, not during one.

• Financial messaging — interbanking transactions, the MT and ISO 20022 standards, and why the unstructured content inside them is the problem TxFlow addresses.
• The case for AI in ISO 20022 — why the richer data foundation matters beyond compliance, and where it supports payment investigations and financial crime work.

Tutorials

End-to-end pipelines on Google Cloud. Part 2 assumes Part 1.

• TxFlow pipeline, part 1 — Dataflow to BigQuery, nested format — a real-time pipeline over SWIFT MT with Pub/Sub, Dataflow and BigQuery, landing the data nested for exploratory analytics.
• TxFlow pipeline, part 2 — Dataflow to BigQuery, tabular format — the same pipeline landing flat, queryable tables instead, for tools that expect SQL rather than nested records.

Conditions

• End user license agreement — the licence agreement between you and Alpina Analytics LLC covering use of TxFlow.
• Terms and conditions — the terms governing use of this website and the services described on it.