What Parabola’s customers and engineers built in July

What our customers built in July 

Inbound PO management Flow

“If you can find a way to cut through the noise and get actionable data in a way that looks nice…the value is tremendous.”

— Director of Logistics at a bedroom furniture company with 50-200 employees

Inbound PO management is a challenge for any brand that needs to track shipments — especially when the process in place involves manual data entry (like adding shipment container numbers and the ultimate container destination) and inconsistent data sources. 

This furniture brand worked with the Parabola team to build a Flow that would:

1. Pull in data from the brand’s two freight forwarders, clean the data so it reflects the same style and insights, then send that data to a Google Sheet 
2. That sheet then becomes a reference that applies logic around how to standardize the shipment status (i.e. if freight forwarder 2 calls the status “shipment in progress,” it will be changed to “en route”)
3. And the output from the Flow is a cleaned version where the brand has data from both freight forwarders now combined, sorted, and categorized the same way
4. It also provides updated dates, which lets them build out a forecast for their team for warehouse receiving 

CI/PL matching 

“It’s exactly what we need.” 

— Senior Director of Logistics at a running gear brand

The team at a running wear brand was struggling to identify and rectify different types of mismatches between their commercial invoices and packing lists. With Parabola, they were able to build a mismatch Flow that doesn’t just identify mismatches, but also drills down to show color, size, and PO mismatches.

While the Flow has a significant amount of steps to handle of the data cleansing and comparison across different potential mismatches, the process ultimately: 

1. Pulls in data from packing lists stored in Excel and commercial invoice PDFs that come in via email
2. Isolates PO #, color, and size within the data to then perform matching comparisons for each of them separately before generating a full mismatch list, in addition to a mismatch list per category
3. Sends the team an email detailing the results of the matching

Now the team has complete visibility into how their packing lists and invoices compare — and it’s all automated. 

What our engineers built in July 

It was a busy month for Parabola engineering, with a focus on launching features and updates that support processing files from URLs, giving feedback when there are email parsing failures, and major updates to the Filter rows and Add if/else column steps. Plus our usual bug stomping.

Pull files into Parabola from URLs

Up first, the team worked on a suite of updates that let users pull files into Parabola from URLs. This means that instead of needing to upload PDFs — or pull in attachment data from emails — users can now use the Pull from file queue source step to access and parse files at the end of URLs. 

The file queue works like a webhook or email trigger but can be triggered by an API call (webhook) or the Run another Flow step.

To make this successful, the team also #shipped:

• New options to run Flows from other Flows
• Updates to email trigger Flow failures
• Pull URLs from emails

Run another Flow

You can now use the updated options in the Run another Flow step to run Flows in sequence. The new options are:

• Run once per row
• Run once per row with a file URL

If Flow 1 tells Flow 2 to run, these options will allow Flow 1 to finish running before Flow 2 has finished. In the existing options, Run once and wait and Run once per row and wait, the step will wait for Flow 2 to finish before it can finish.

Using the Run once per row with a file URL option will add runs to the file queue of Flow 2

Pull URLs from emails

We’ve updated the Pull from inbound email step with options to pull in the full HTML representation of the body of an email, as well as a parsed list of URLs contained within the body.

This list of URLs can be used to isolate URLs from an email, fetch a document from that URL, and then process that document via the Pull from file queue step.

Updates to email trigger Flow failures

Parabola Flows can be set up to be triggered to run via an email. When an email is sent to a Parabola Flow, the Flow will try to enqueue the new run. If that enqueueing fails, an email is sent back to editors of the Flow alerting them that a certain email could not be processed.

With this change, that email also:

• Contains the name of the file attached that caused the issue (such as the wrong file type)
• Sends to the email address who sent the original email in to Parabola

Major updates to Filter rows and Add if/else column steps

Filtering rows and if else logic are at the heart of every Parabola Flow. We’ve given both steps a massive upgrade to make them faster, easier to use, and more powerful

Both steps have access to series of new operations:

• Filter dates to… (filter dates relative to now or a set date)
• Is unique (is the value in each row unique within that column)
• is not unique (is the value in each row not unique within that column)
• Text is equal to (equals but assumes text - faster than “is equal to”)
• Text is not equal to (equals but assumes text - faster than “is not equal to”)
• Text starts with (matches the first part of a cell)
• Text ends with (matches the last part of a cell)
• Text length is (length of cell = #)
• Text length is greater than (length of cell > #)
• Text length is less than (length of cell < #)
• Text matches pattern (regex match)
• Text does not match pattern (inverted regex match)
• Is between (between two numbers)
• Is not between (not between two numbers)

And some new text field level options, which are accessible via the settings cog icon on the right side of any text field:

• Each text field can be toggled to evaluate what has typed in as a math expression, instead of just text. You can type numbers or reference them from columns using {merge tags}
• Each text field can be toggled to match based on the casing provided, or ignore the casing of the text

Add if/else column step-specific updates:

• Choose to replace value in an existing column, as opposed to only create values in a new column
• Add additional criteria to a single condition, which was already supported in the Filter rows step. For example, you can define a list of items that a field may be equal to.
• To accommodate all of these features, the Add if/else column step also has a new design that is more compact and easier to read

Bug fixes 🐞

Plus, a couple of pesky bugs have been addressed. Here’s what you can look forward to:

1. Updates to the Send to Salesforce step that would sometimes randomly error or timeout. It’s now much faster and more reliable. 
2. If you trigger any Flows from a web button or URL, you may have been running Flows twice due to browser optimizations that start processing the link before you actually click on it. We’ve set up guardrails so that the Flow only runs when you click the link or the button now. 
3. Case sensitivity was impacting the reliability of Parabola pulling in files from emails (i.e. if the Flow was looking for .pdf but found .PDF if might not pick it up). We’ve fixed this.
4. Maybe you’ve noticed a purple “beanie” on the product UI for any of our AI-powered steps so you can start to see when you’re using AI, especially as we roll out more new features.
   

Mark your calendars…

The SVP of Supply Chain and Quality at Seed Health is hosting an AMA for The SOP Community. 

Parabola’s Head of Customer Experience is going deep on how to make your Shopify data usable with Parabola.

Back to Parabola basics — if you want to learn how to clean, combine, and transform data with Parabola, this one is for you.

If you’re in NYC on August 28th, we’re hosting a dinner for ops leaders. Respond to this email if you’d like to learn more.