ingestr v1.0.44 - Streaming Sources and Braze Series Tables

ingestr v1.0.44 was published on June 24, 2026, with streaming platform sources as the main change for ETL and ELT teams. The release also broadens Braze extraction with additional series tables, which matters for teams that model campaign and canvas behavior outside the source system.

The full release notes and downloads are on the GitHub release page.

The first changelog item is concise: add streaming platform sources. The release notes do not list the individual platforms, so the practical reading is cautious. Operators should treat this as a source surface expansion and verify connector coverage against their own source catalog before changing pipeline contracts.

For data teams, the interesting part is less the label “streaming” and more the way it changes ingestion planning. A source that represents streaming platform data often has time windows, event identity, delivery delay, and replay behavior that differ from a simple SaaS object table. Even when a connector exposes the data through a batch extract path, the downstream model usually needs clear handling for late events and partial windows.

This release also updates licenses.lock.yml and regenerates THIRD_PARTY_LICENSES.txt for the streaming source dependencies. That is not a runtime feature, but it is relevant for platform operators. New connectors often bring new client libraries, compression libraries, and transitive dependencies. Keeping the license manifest in the same release reduces audit drift.

The Braze changes are more explicit. ingestr v1.0.44 adds campaign_series and canvas_series, then adds sessions, purchase, and segment series tables. That gives analytics teams more source level structure for modeling Braze activity without flattening every report into a single campaign view.

The useful angle here is granularity. Campaign and canvas level tables are good for top level reporting, but series tables are where lifecycle and journey analysis often becomes easier to model. If a warehouse project already has Braze facts joined to campaigns, these new tables may allow cleaner dimensions for series behavior and less custom stitching in dbt or SQL transforms.

There is also a small hygiene change: Braze series comments are trimmed to at most 2 lines. That sounds cosmetic until those comments land in generated schema docs, catalog metadata, or exported table descriptions. Long source comments can make catalogs noisy, especially when comments are copied across environments.

One changelog item addresses skipped segments with a non-string id. The important detail is that skipped segments are now logged, rather than silently hiding an awkward source record shape. For operational ETL, that distinction matters.

Source APIs do not always keep identifiers as clean strings. A segment object with an unexpected id type can break assumptions in extract code, schema inference, or error reporting. Logging skipped records gives teams a path to answer a basic production question: did the job succeed because all data matched expectations, or did it succeed while discarding records that should be investigated?

This is not a substitute for a strict data quality check downstream. It does make the first diagnosis easier. When a pipeline owner sees a missing segment in the warehouse, the extract logs should now have a better chance of showing whether ingestr skipped a record because the id did not match the expected type.

The remaining changes are mostly project maintenance. The release adds DataDog/zstd to licenses.lock.yml, fixes a license test, and refreshes third party license output. These are not user visible features, but they are part of keeping an ingestion tool usable in controlled environments.

The release notes do not mention breaking changes or migration steps. They also do not mark v1.0.44 as a prerelease. For a platform team already running ingestr, the likely upgrade review is connector specific: check whether the new streaming sources or Braze tables affect dependency approval, warehouse naming conventions, and table allow lists.

• Release page: GitHub release page
• Project repo: bruin-data/ingestr
• Tag: v1.0.44