Lens : See what Adobe tools don't show you [Source DataFlow]

ADOBE EXPERIENCE PLATFORM (AEP)

Pradeep Jaiswal

8/10/20266 min read

Data enters Adobe Experience Platform from dozens of sources: cloud storage buckets, CRM systems, data warehouses, analytics report suites, streaming inlets, and Query Service jobs that write derived tables back into the platform. Each ingestion flow passes through a chain of platform objects ; a source connection, a flow configuration, a target connection, a dataset ; before the data arrives at a schema where it can be queried and activated.

This chain is not visible anywhere in the AEP interface as a whole. You can navigate to a specific flow and see its configuration. You can look at a dataset and trace back to its schema. But seeing all of your ingestion architecture in one diagram, understanding which source systems feed which schemas, and identifying gaps in coverage ; that requires a tool that is not built into AEP.

The Source Dataflow Canvas in Lens is that tool.

Chrome extension. Link

The canvas renders every source ingestion flow in your active AEP sandbox as a connected chain of nodes, one chain per flow. A complete chain has five segments:

Source System → Source Location → Flow → Schema → Dataset

Source System: This is the origin of the data. Lens resolves the human-readable name and logo for every source connector type that AEP supports — Amazon S3, Snowflake, Adobe Analytics, Google Pub/Sub, Salesforce, MySQL, and dozens more. Where a logo is available, it appears inside the node. Where no logo exists, a colored letter-avatar using the source name's initial is shown.

The source system node answers: "Who made this data?"

Source Location: This is the specific resource within the source system: the S3 bucket path, the Snowflake table name, the Analytics report suite ID, the Pub/Sub topic, the PostgreSQL table, the SFTP path. Different source systems expose this in different ways; Lens normalizes them into a single "location" label that makes sense for each connector type.

The source location answers: "Exactly where in that system does this data come from?"

Flow: The flow is the AEP ingestion configuration that defines how data moves from the source to the platform. It carries the flow name, the run state (enabled, disabled), and the schedule frequency (hourly, daily, weekly, streaming, ad-hoc). A disabled flow is shown with reduced opacity and a grey status indicator, visually distinct from an active one.

The flow node answers: "Is this pipeline running, and how often?"

Schema: This is the AEP schema that the flow's data is mapped to. It is the bridge between the raw source data and the structured, queryable data model inside the platform. A schema here has the same identity as a schema you would see in the ERD Canvas ; it is the same object, just viewed from the data movement angle rather than the structural relationship angle.

The schema answers: "What shape does this data take inside the platform?"

Dataset: The dataset is the physical storage partition where records for this flow are written. In AEP, every dataset corresponds to a physical table in the underlying Query Service (PQS). Lens shows the dataset name and its physical table name ; useful when running queries in AEP Query Service or external database and needing to know the exact table to reference.

The dataset answers: "Where is the data actually stored?"

Try Lens

Three Ways to Group the Canvas

The canvas has three grouping modes, switchable from the toolbar. Every mode draws the same five-column chain; what changes is which column holds the anchor that everything else clusters around.

Dataset View: Dataflows feeding a specific dataset. Each dataset becomes a hub, and every flow writing into it converges there. This is the physical-storage view: the question it answers is "what is writing into this dataset, and from where?"

Schema View: Dataflows sharing a specific schema. Each schema becomes a hub. If your Customer Events schema receives data from three different analytics report suites and a streaming inlet, you see four chains converging into one schema box. This is the data-model view, most useful when the question is "what sources feed this schema?" and you want all of them at once. Note that schema and dataset views often look similar, because most flows map one schema to one dataset. The difference becomes informative exactly when that assumption breaks.

System View: Dataflows originating from a specific source system. Each source system becomes the anchor, and every flow it feeds fans out to the right. This is the vendor view: "everything Snowflake sends us," "everything arriving from Analytics." It is the fastest way to scope the blast radius of a source system going offline or a credential rotating.

End-to-End Line Selection

When a schema or dataset receives data from many flows, the canvas can become dense. Fifteen lines entering a single schema node are hard to trace individually. Click any line on the canvas. Every segment in that flow's four-connection chain ; source-to-location, location-to-flow, flow-to-schema, schema-to-dataset ; highlights in full color. Every other line dims to a faint grey. The result is a single highlighted path from source system to physical dataset, with everything else pushed to the background. Click the same line again, or click on an empty area of the canvas, to clear the selection.

Exporting the Diagram

Three exports, depending on where the picture needs to end up.

CSV for the table, as it be used as the input to a spreadsheet, a BI tool, or a governance document.

Draw.io / Lucidchart for the diagram. The export writes a real `.drawio` file with the nodes, the routed connections and the grouping preserved, so it opens as an editable diagram rather than a flat image. This is the one to use when the architecture is going into a slide deck or a document that someone will annotate.

PNG for a high-resolution capture of the whole canvas, including the parts scrolled off screen.

Schema View

Dataset View

System View

Practical Use Cases

Ingestion coverage audit. Open the Schema View and look for grey "Not Connected" schemas. Every grey box is a schema in your data model with no active data source ; the fastest way to find schemas that were created and never wired up.

Flow health check. In the Table View, filter State to `disabled`. Every row is a pipeline that has been turned off. When an analyst reports missing data for a source, this is the first place to look.

Blast radius of a source system. Switch to System View and find the vendor. Everything fanning out to the right depends on that connection. Before a credential rotation or a connector migration, this is the list of things that can break.

Query Service lineage. Query Service flows show which derived tables are built from which source datasets. When a downstream aggregate looks wrong, this traces it back to its inputs without reading through the SQL.

Source-to-schema documentation. Export the table to CSV, or the diagram to draw.io, and hand it to your governance team. Between them you have a complete inventory of every source, its frequency, its schema and its dataset.

New flow validation. After configuring a source connection and flow in AEP, click Refresh. The flow should appear in both canvas and table. If it appears with "Not Connected" placeholders, the configuration has a gap ; the schema reference or dataset assignment is incomplete.

Multi-source schema impact analysis. In Schema View, count the lines entering the schema you plan to change. Each is an independent source. Adding a required field or changing an identity field must be validated against all of them, and the canvas makes that scope visible before you start.

Table View

Need Expert Help Mastering Your CDP Strategy?

At Shiftlytic, we specialize in Customer Data Platform (CDP) architecture, helping brands like yours consolidate fragmented data from website interactions, CRM systems, and offline sources into a unified 360-degree customer profile. We focus on turning dormant data into actionable insights that fuel real-time personalization and seamless omnichannel journeys.

Don’t let the technical complexity of cloud migration, the limitations of legacy batch processing, or the overhead of duplicate data storage stall your innovation. Whether you are exploring the modular flexibility of a Composable CDP integrated with your existing warehouse or implementing a full-stack solution like Adobe Experience Platform, we ensure your data integrity is maintained through robust cleansing and normalization.

Contact us today to schedule one free consultation and ensure your CDP ecosystem is accurate, scalable, and fully optimized. Let’s turn your customer data into a competitive advantage together