End to end pipeline

Understanding the end-to-end workflow

Before defining solutions, I mapped the complete translation workflow across translators, reviewers, and project managers to understand where information, ownership, and context were lost. Rather than optimizing isolated screens, this systems view allowed me to identify recurring friction across the entire pipeline and prioritize the problems with the greatest impact on collaboration and delivery.
PM
Client request becomes a Catch quote & deadline
→
PM
Offer sent to translator & proofreader by score
→
Translator
NMT pre-translates, linguist reviews & confirms segments
→
Proofreader
Consistency & fluency review, terminology confirmed
→
PM
QA, spot check & delivery to client

Translator journey

13 steps · 4 pain points, as mapped in research
  • 1Get the offer in the linguist portal
  • 2Check the deadline and word count
  • 3Accept the job by email
  • 4Review job specifications and reference material
  • 5Open the project in Catch with an emailed linkPain
  • 6Set up the workspace, source file next to the editor
  • 7Translate the segments and confirm each onePain
  • 8Finish translating, check terminology, fluidity & localization
  • 9Run QA
  • 10Download the file and compare formats against the originalPain
  • 11Click finish and fill the feedback popup
  • 12Go back to the linguist portal and refresh the link
  • 13Confirm the status moved from "running" to "delivered"Pain
✓4 of 13 steps were pain points. Translation Memory and Segment History removed the guessing while confirming segments and the portal refreshes to check delivery status.

Proofreader journey

13 steps · 4 pain points, as mapped in research
  • 1Get the offer in the linguist portal
  • 2Check the deadline and word count
  • 3Accept the job by email
  • 4Review job specifications and reference material
  • 5Open the project in Catch with an emailed linkPain
  • 6Set up the workspace, source file next to the editor
  • 7Proofread the segments and confirm each onePain
  • 8Finish proofreading, confirm terminology and localization
  • 9Run QA
  • 10Download the file and compare formats against the originalPain
  • 11Click finish and fill the feedback popup
  • 12Go back to the linguist portal and refresh the link
  • 13Confirm the status moved from "running" to "delivered"Pain
✓4 of 13 steps were pain points. Comments replaced the email back and forth with translators, Segment History made every edit traceable without leaving the tool.

Project manager journey

18 steps · 7 pain points, as mapped in research
  • 1Receive the request from the client
  • 2Convert the request to an order in the admin portal
  • 3Get the client's approval on the order
  • 4Create the Catch quote
  • 5Send the offer to multiple translators & proofreaders by score and industryPain
  • 6Set the delivery deadline to the client's wishes
  • 7Check which translator and proofreader accepted
  • 8Wait for the translation to be readyPain
  • 9Receive the project from the proofreader or translator
  • 10Do final checks on segments linguists had questions onPain
  • 11Run QA
  • 12Download the original and translated files from Catch
  • 13Review and compare the files in WordPain
  • 14Anonymise the files so the client never sees author names
  • 15Rename the files back to the original namesPain
  • 16Replace the files with the final ones in the portalPain
  • 17Email the client that the project is finishedPain
  • 18Close the project in the admin
✓7 of 18 steps were pain points. QA replaced the manual Word file comparison, and live progress tracking removed the waiting and file juggling before delivery.
Translator Proofreader Project manager Select a role to explore its journey