> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.fynapse.app/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.fynapse.app/_mcp/server.

> Fynapse releases enhancements on the Adjustment Portal screen, with new functionalities, such as error handling, and redesigns the Processing Summary screen and adds a Pipeline view that follows each ingestion through every stage of its processing, Flows included.

## Processing Summary Overhaul

processing-summary, flows, ui migration

We redesigned the **Processing Summary** screen and added a second view, **Pipeline**, that follows each ingestion through the stages it actually passes, including the Flows that read its records.

The changes are listed below.

> **Info**
>
> The classic view keeps its columns, statuses and counts.

### Changes on the Classic View:

* **Refresh** is now in the page header. The **Classic** | **Pipeline** switch, **From** date box and **Reset** are in a toolbar above the grid.
* The screen now refreshes every 30 seconds. You can no longer choose a refresh interval.
* **From** is set to midnight seven days ago by default, and **Reset** restores that value.
* The grid now shows 100 ingestions per page with a pager, instead of the **Load more** button.
* Ingestions are now sorted by start time, newest first.
* An opened ingestion now shows its counts in a **Subledger statistics** section, with **Export to Excel** beside the heading.
* A count that is too wide for its column now ends in an ellipsis. Widen the column to see the whole number.

### New Pipeline View:

* Each ingestion has a strip of the stages it passed, such as **Ingested**, **Validated**, **Flows**, **Accounting**, **Posted** and **Stored**. Each stage shows the records that passed it and the records that did not: duplicates, failures, records a Flow produced no output for, and unposted Journal Lines.
* The **Kind** column shows whether the ingestion contains Business Events, Journal input, reference data or transaction data, its Entity, and whether it came from a file, the API or an extract.
* The Pipeline view also lists the ingestions that the Classic view does not: reference data, and ingestions read by Flows that write no Business Events or Journal input.
* You can filter the grid by ingestion, kind and status.
* Click an ingestion to open its page: a diagram of its stages, a table of the Flow versions that read its records, and its Subledger statistics. **N failed records** opens the **Errors** screen filtered to the ingestion.

> **Info**
>
> The Pipeline view lists ingestions that start after your environment is upgraded to this release. Earlier ingestions remain in the Classic view, which lists the ingestions that lead to Journals.

Read the docs

Video preview:

## Adjustment Portal Enhancement

adjustment portal

We introduced a number of enhancements on the **Adjustment Portal** screen.

Previously, when uploading a file with Journals, there were two ingestion error scenarios:

* The file had invalid format or header - the file would receive **Invalid upload** status, no records were ingested and you had to download the errors in a .CSV file
* The records in the file were invalid - the file would receive **Invalid upload** status, the records were ingested but failed processing; you had to download the errors in a .CSV file

### Changes

* Error handling is now available directly on the **Adjustment Portal** screen.
* We introduced a new status: **\[number of] errors** for files with errors that can be resolved in the UI
* Additionally, the errors are available as tooltips.

So now, the two scenarios have different outcomes:

* When a file fails upon ingestion, e.g. has invalid format or header - the file still receives **Invalid upload** status and no records were ingested
  but instead of downloading the file, you view the errors on the **Details** screen for the given ingestion. These types of errors
  require changes to the file, e.g. reformatting or amending the header, and cannot be resolved on the screen.
* When a file is ingested, but the records in the file are invalid, the file receives the new status **\[number of] errors**
  in the file and you can view and resolve the errors on the **Details** screen for the given ingestion.

Read the docs

Video preview: