Data Quality & Governance Blog | Insights from Tale of Data

Documenting a data flow in Tale of Data: greater clarity, less loss of context

Written by Adnan Joudeh | (July 2025)

Documenting a data flow in Tale of Data: greater clarity, less loss of context

Between business logic and technical processing, a data flow deserves more than a simple diagram. In this article, you'll discover how Tale of Data lets you add explanations directly to the design canvas. A new collaborative feature designed to document where it all happens, without external tools, without overload, and above all without loss of meaning.

What are the concrete benefits of documenting a data flow?

A process may be technically sound, but difficult to understand. This lack of explanation slows down projects far more often than you might think. When a flow is designed, everything seems clear at the time. But a few days later, or when another user has to take over, the intentions become blurred. Why this filter? What business logic justifies this transformation? What assumptions guided this choice?

All too often, documentation is absent, scattered or disconnected from the tool. A forgotten file, a note in a ticket, a few exchanges in a meeting. Nothing lasting. As a result, everyone interprets things in their own way, errors multiply and delays lengthen. And when there's a lack of shared vocabulary between technical and business profiles, understanding becomes even more fragile.

Tale of Data has introduced a new feature in its Flow Designer to meet the dual challenge of making processing easier to understand and aligning the players involved. It enables you to annotate each stage of a flow directly on the canvas, where decisions are visible and usable.

This approach benefits all profiles:
- Developers find it a reliable maintenance tool.
- A business user can explain a rule or objective.
- A CDO sees it as a governance lever.
- A CISO reads it as a transparency signal.

It proves particularly useful in the context of collaborative projects, regulatory compliance (RGPD, BCBS 239...), the resumption of old processing or in-depth business documentation. It provides a direct response to a universal need: to understand what has been done, how and why.

And to take this alignment logic a step further, the language of data quality, provides a foundation of shared definitions to facilitate exchanges between all data stakeholders.

Documentation directly integrated into the interface

There's no need to use a separate tool or external document: each step in a flow can now be commented on directly in the Tale of Data Flow Designer. It is possible to add an overall description to the flow, but also to annotate each node individually, in the form of post-its anchored to the canvas, which can be activated or hidden as required.

The integrated editor allows you to explain a rule, justify a choice or contextualize a treatment, with no format constraints. You write simply, as you would for a colleague. No overload, no detours. It's not cosmetic, but a way of embedding business logic where it's used, in a durable and legible way.

Activation is immediate: description display is active by default for all new flows. For older flows, a simple click in the side menu is all that's required.

Each node can be annotated at any time. Text can be free-form or structured, as required. No learning curve, no parameters to set. The interface remains fluid, even when enriched.

An annotated flow: an example applied to the detection of temporal duplicates

The above flow is used to identify temporal duplicates in data from industrial sensors. This type of anomaly occurs when several measurements are recorded at exactly the same time for the same sensor, an unusual situation which can distort analyses.

Each processing step is enriched by a visible annotation on the canvas, describing its logic and purpose. This enables any user, even those who did not create the flow, to understand the entire reasoning in a matter of moments.

The pipeline runs as follows:

  • As input, a file structured with Tag, Time and Value: sensor name, measurement timestamp, recorded value.

  • The data is aggregated by timestamp to identify instants present several times, indicating a duplicate.

  • Filtering eliminates unique measurements and retains those that share the same instant.

  • Duplicates are isolated in a dedicated output for analysis or correction.

  • A global note explains the logic of the flow and its business purpose, directly in the interface.

This type of integrated documentation avoids haphazard interpretations, facilitates re-reading, and strengthens collaboration between technical and business profiles. You no longer read a process: you follow a clearly explained logic.

What this means in concrete terms

Documenting a data flow where it is created transforms the way we work with data.

We no longer depend on a Word file or a forgotten slide. Documentation becomes an integral part of the pipeline. It's always up to date, always visible, always associated with processing.

This facilitates transmission, avoids misunderstandings, speeds up project rework and limits errors. Above all, it gives the whole organization greater confidence in the quality of workflows.

To discover this feature

This new feature is already available in all Tale of Data environments.

It fits naturally into your day-to-day use, without disrupting your practices. If you're already using the Flow Designer, it's ready to use. And if you're new to Tale of Data, it's a great opportunity to rethink the way you document and govern your data pipelines.