Platform / Data in

Listen

Data enters here. Workflows and agents point at it. They do not carry the password.

Product screens recreated with synthetic data.

Listen · ap-inbox
Datasets · AP workspace
ap-inbox412 files · mounted read-only
vendor-masterlive connection · synced
Files
INV-20841.pdf184 KBSep 26 09:14
INV-20840.pdf96 KBSep 26 09:02
manifest.json2 KBSep 26 08:40
Preview · INV-20841.pdf
Vendor   Acme Industrial
Invoice  INV-20841
Amount   $12,480.00
Due      Oct 14, 2026

The invoice pipeline reads this dataset by reference.
The key that lists the bucket is not in the block.

Pipelines, workloads, and dev spaces mount this source. They do not get a copy of the credential.

What it is

Listen is the front door for data. A dataset you upload, a bucket, a connection you keep in sync — each one is registered on the workspace, with an owner and a status you can see. nQube blocks and workloads mount that source. They do not get a fresh copy of the credential pasted into a prompt.

Freshness is part of the screen. A source that is syncing, and a source that last synced yesterday, look different before a workflow fails on stale files. The file browser lists what is there, previews text, and can import it, so “what did we actually load?” has an answer in the console.

What it is not

Listen does not send alerts and it does not post results. Outbound is a declared capability on the workflow — a webhook commit under agentOS, or a destination you wire there. Listen is inbound only.

Who it is for

Whoever is tired of credentials living in workflow configs, and whoever has to prove which copy of a file a run actually read.

Where you get it

Platform. Datasets are on Team. The source stays in your workspace; pipelines, workloads, and dev spaces refer to it.

Pricing →

The invoice folder

Synthetic example. Not a customer.

You register the bucket where AP drops PDFs. The invoice pipeline’s dataset reader points at that registration. A new batch shows up in the browser with sizes and times. The workflow run reads that source. The model never sees the key that lists the bucket.

What changes when you have it

One registration

The same dataset feeds a pipeline, a workload, and a dev space.

Status you can see

Syncing versus synced, before a downstream run surprises you.

A browser, not a ticket

List, preview, and import text files in the console.

Inside Listen

  • Datasets and connections per workspace
  • File browser with size and modified time
  • Inline preview and text import
  • Sync status on each connection
  • Mounted by pipelines, workloads, and dev spaces

How it works

1

Register the source

A dataset or a live connection, on a workspace, with an owner.

2

See if it is fresh

Sync status is on the source. Stale is visible here.

3

Point the work at it

nQube readers and mounted workloads use the registration. No ad-hoc secrets in the graph.

4

Inspect the files

Listing, size, modified time, inline preview, import for text.

Listen. Data in through Listen. A write out is a capability on the workflow, enforced by agentOS.

The names next to it

See it on your workflow

Platform Start is the self-serve canvas. A first pipeline we stand up with you is delivery, priced apart from the subscription.