Datasources

A datasource is a node that brings information into a pipeline from a system outside Quantra — a network folder, a mailbox, a document library, a database, and so on. A pipeline normally starts with one or more datasource nodes. This section describes the datasources you can choose from, what each one does, and what to fill in when configuring it.

Box

Connects to a Box account and brings documents from your Box folders into the pipeline. Use it whenever the documents you need to process live in Box, whether they are uploaded by colleagues, shared in by external partners, or stored in long-term Box archives.

What you'll need

• A Box account with access to the folders you want to read.
• An Open Authorization (OAuth) application registered in Box for Quantra. This is normally already in place — see the administrator note below.

Settings you provide

Setting	Required	Description
Display name	No	A label for this node on the canvas, for example "Customer uploads".
Authenticate with Box	Yes	Click the button to sign in to Box and authorise Quantra to read your files. After successful sign-in, the connection is remembered for this node.

Common uses

• Processing documents that external parties upload to a shared Box folder.
• Bulk-processing a Box folder of historical documents.
• Triggering a workflow whenever a new document appears in a Box folder.

M-Files

Connects to an M-Files vault and brings documents and their associated metadata into the pipeline. Use it when your team uses M-Files as the system of record for contracts, policies, case files, or other controlled documents.

What you'll need

• The web address (URL) of your M-Files server.
• The vault's Globally Unique Identifier (GUID), which looks like {XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX}.
• Either an M-Files username and password, or a sign-in token issued by your M-Files administrator.

Settings you provide

Setting	Required	Description
Display name	No	A label for this node on the canvas.
Server URL	Yes	The web address of the M-Files server, for example https://mfiles.example.com.
Vault GUID	Yes	The unique identifier of the M-Files vault you want to read from.
Username	One of these is required	Your M-Files username, in either DOMAIN\user or user form.
Password	One of these is required	Your M-Files password.
Authentication token	One of these is required	An M-Files sign-in token if your administrator has issued one. Use either username and password or a token, not both.

Common uses

• Running compliance checks across documents in an M-Files vault.
• Extracting and reusing metadata held against M-Files documents.

Local Drive

Reads files from a specific computer that has the Quantra Local Agent installed on it. The agent is a small program that runs on a workstation or server and exposes a chosen set of folders to Quantra over a secure connection. Files never leave the host machine until you explicitly fetch them through a pipeline. Use this datasource when the files you need to process sit on a personal or office machine rather than in a shared system.

What you'll need

• A computer that has had the Quantra Local Agent installed and enrolled by an administrator.

Settings you provide

Setting	Required	Description
Display name	No	A label for this node on the canvas, for example "Office laptop".
Agent	Yes	Pick the enrolled machine you want to read from. The dropdown lists every agent your administrator has enrolled and that you have access to.

Common uses

• Processing files from a personal workstation without first uploading them anywhere.
• Pulling files from a machine that is not exposed on the network as a shared drive.

Microsoft SQL Server

Connects to a Microsoft SQL Server database. You can read records from existing tables to feed a pipeline, and you can also write records produced by a pipeline back to a chosen table. Structured Query Language (SQL) is the standard language used to read from and write to such databases.

What you'll need

• The server name (or address) of the SQL Server instance you want to use.
• A database username and password with the right level of access to the tables you need.

Settings you provide

Setting	Required	Description
Display name	No	A label for this node on the canvas.
Server	Yes	The hostname or address of the SQL Server. If your administrator has pre-registered approved servers, you can pick one from the dropdown instead of typing the address.
Username	Yes	Your database username.
Password	Yes	Your database password.
Sample rows	No	How many rows to show in previews while you design the pipeline. Defaults to 5.
Write mode	No	When the node is being used to write records, choose Insert (add new rows) or Upsert (insert new rows or update existing matching rows). Defaults to Insert.
Match keys	Only for Upsert	The column names that identify a row uniquely, separated by commas. Used to decide whether to insert or update.
Create table if missing	No	Automatically create the target table if it does not already exist. Defaults to on.
Allow new columns	No	Automatically add new columns to the target table when the pipeline produces fields that are not yet there. Defaults to on.
Overwrite existing	No	Replace the contents of the target table before writing.

Common uses

• Pulling records out of a line-of-business system to enrich documents with metadata.
• Writing the results of a pipeline (for example, extracted fields from invoices) back into a SQL Server table.

Outlook

Connects to a Microsoft 365 mailbox and brings emails — and their attachments — into the pipeline. Use it whenever your process starts with messages arriving by email, such as application forms, supplier invoices, or customer correspondence.

What you'll need

• Access to the mailbox you want to read.
• A registered application in Microsoft Azure Active Directory (Azure AD) that allows Quantra to read mail. This is normally already in place — see the administrator note below.

Settings you provide

Setting	Required	Description
Display name	No	A label for this node on the canvas.
Authenticate with Outlook	Yes	Click the button to sign in with your Microsoft 365 account and authorise Quantra to read your mail.
Mailbox	No	The email address of the mailbox to read. Leave blank to use the mailbox of the account you signed in with.
Folder	No	The mail folder to read from, for example Inbox or Inbox/Submissions. Defaults to Inbox.
Maximum items	No	The largest number of messages to read in one run. Defaults to 100.
Days to look back	No	How many days of mail to consider. Defaults to 30.

Common uses

• Processing email-based document submissions such as invoices and applications.
• Extracting and archiving email attachments for compliance.
• Triaging and routing incoming correspondence automatically.

PostgreSQL

Connects to a PostgreSQL database. As with the Microsoft SQL Server datasource, you can read records to feed a pipeline and write records produced by a pipeline back to a chosen table.

What you'll need

• The server name (or address) of the PostgreSQL instance you want to use.
• A database username and password with the right level of access.

Settings you provide

Setting	Required	Description
Display name	No	A label for this node on the canvas.
Server	Yes	The hostname or address of the PostgreSQL server. If your administrator has pre-registered approved servers, you can pick one from the dropdown instead of typing the address.
Username	Yes	Your database username.
Password	Yes	Your database password.
Sample rows	No	How many rows to show in previews while you design the pipeline. Defaults to 5.
Write mode	No	Insert (add new rows) or Upsert (insert new rows or update existing matching rows). Defaults to Insert.
Match keys	Only for Upsert	The column names that identify a row uniquely, separated by commas.
Create table if missing	No	Automatically create the target table if it does not already exist. Defaults to on.
Allow new columns	No	Automatically add new columns to the target table when the pipeline produces fields that are not yet there. Defaults to on.
Overwrite existing	No	Replace the contents of the target table before writing.

Common uses

• Reading from open-source application databases that use PostgreSQL.
• Writing pipeline results into an analytics or reporting database.

SharePoint

Connects to a Microsoft SharePoint site and brings documents from a chosen document library into the pipeline. Use it when your organisation uses SharePoint to store policies, contracts, project documents, or any other shared content.

What you'll need

• Access to the SharePoint site and document library you want to read from.
• A SharePoint or Microsoft 365 username and password.

Settings you provide

Setting	Required	Description
Display name	No	A label for this node on the canvas.
Username	Yes (unless a group credential is in place)	Your SharePoint or Microsoft 365 sign-in name, for example user@company.com.
Site	Yes	The URL of the SharePoint site, for example https://tenant.sharepoint.com/sites/Project.
Library	Yes	The name of the document library to read, for example Documents.
Root folder	No	A subfolder within the library to start from. Defaults to the top of the library.
Recurse into subfolders	No	Read documents from subfolders too, not just the root folder.
Include earlier versions	No	Also bring in previous versions of each document, not just the latest.

Common uses

• Running classification, redaction, or summarisation across a SharePoint document library.
• Migrating documents out of SharePoint while transforming them in the pipeline.

Snowflake

Connects to a Snowflake cloud data warehouse. As with the other database datasources, you can read records into a pipeline and write pipeline results back into a Snowflake table.

What you'll need

• Your Snowflake account identifier (for example, myorg-myacct).
• A Snowflake username and password.
• The names of the warehouse, database, and schema you want to use, if these are not already set as defaults for your account.

Settings you provide

Setting	Required	Description
Display name	No	A label for this node on the canvas.
Account	Yes	Your Snowflake account identifier.
Username	Yes (unless a group credential is in place)	Your Snowflake username.
Password	Yes (unless a group credential is in place)	Your Snowflake password.
Role	No	The Snowflake role to use, for example SYSADMIN. Leave blank to use your default role.
Warehouse	No	The Snowflake virtual warehouse to use for queries.
Database	No	The Snowflake database to read from or write to.
Schema	No	The schema within the database.
Sample rows	No	How many rows to show in previews while you design the pipeline. Defaults to 5.
Create table if missing	No	Automatically create the target table when writing if it does not already exist.

Common uses

• Pulling reference data from an enterprise data warehouse to enrich documents.
• Writing structured pipeline output (extracted fields, scores, classifications) into Snowflake for reporting.

Network Drives

Connects to a network file share — the kind of shared folder you might map as a drive letter on Windows. Network shares use a protocol called Server Message Block (SMB), which is the same protocol Windows itself uses for file sharing. The datasource walks through the chosen folder and brings every file it finds into the pipeline. Use it whenever the documents you need to process live on a shared drive.

What you'll need

• The address of the file server (a hostname such as fileserver01.corp.local, or an address such as 192.168.1.253).
• A username and password with permission to read the share, plus the Windows domain if your environment uses one.

Settings you provide

Setting	Required	Description
Display name	No	A label for this node on the canvas.
Server	Yes	The hostname or address of the file server. If your administrator has pre-registered approved servers, you can pick one from the dropdown instead of typing it.
Username	Yes	Your sign-in name, in either DOMAIN\user or user form.
Password	Yes	Your password.
Domain	No	The Windows domain, if you have not included it in the username field.

Common uses

• Processing batches of scanned documents from a department file share.
• Reading documents from a shared drive for compliance or migration projects.
• Pulling files from an archive server for digitisation.