Helping translators process refugee documents faster with human-reviewed AI drafts
FirstPass AI helped translators move from blank-page document translation to AI-assisted first drafts. I designed the workflow around AI draft generation, human review, editing and final approval so translators could work faster while keeping control, context and accuracy. The work helped reduce average document translation time from 37 minutes 54 seconds to 4 minutes 50 seconds, an improvement of approximately 87%.
Measured outcomes
How the workflow works
The system helps translators start from a structured AI-assisted draft instead of a blank page, reducing the cognitive effort required to begin translation.
The workflow makes it explicit that the translator must review and correct the AI draft before submission — the product positions AI as an assistant, not a decision-maker.
Translators can edit extracted text, review the translated output and correct sections requiring human judgment before the document moves forward.
If a translator tries to submit without making any edits, the product prompts them to review before sending — reducing blind submission of AI-generated output.
Submission feels intentional and human-controlled. The final step requires an explicit action, making responsibility clear and the process trustworthy.
Overview
Tarjimly supports multilingual communication for refugees, nonprofits, hospitals and humanitarian organizations. Many document translation requests involve sensitive materials — medical letters, legal forms, housing documents, asylum records and personal identification.
Before FirstPass, translators often had to review each document manually from the beginning. This slowed down urgent cases and created pressure for teams handling high volumes of sensitive requests.
FirstPass was designed to help teams understand documents faster while keeping human review where it mattered most.
The challenge
The challenge was not simply adding AI to translation. The product needed to help translators move faster without pretending every document could be handled automatically.
The workflow had to support unclear uploads, sensitive content, AI uncertainty, review points and final human approval.
Initial findings
Translators needed speed, but not at the cost of accuracy.
AI output had to be framed as a draft, not a final answer.
Sensitive documents required human judgment.
Users needed clear states for extracted text, translated text, edits and final approval.
The product needed to discourage blind submission of AI-generated output.
Design goals
Reduce document processing time.
Keep human judgment central to the workflow.
Make AI-generated drafts easy to review and edit.
Make responsibility and approval clear.
Build trust in AI-assisted translation without overstating AI accuracy.
Design process
The process included an audit of the existing translation workflow, followed by AI-human workflow mapping to define where the system should assist and where humans must remain in control.
From there I worked through draft generation flows, review and editing states, warning and confirmation states, prototype reviews with product and engineering, and finally handoff and QA.
Designing with constraints
Unclear scans and low-quality uploads
Sensitive refugee documents
Multilingual content
AI uncertainty and confidence gaps
Need for human review before approval
Translation accuracy and cultural context
Mobile-first translator experience
Reusable patterns
The components and interaction states built for FirstPass were designed to be reusable across other AI-assisted workflows — not just document translation.
Launch and quality
The workflow was tested across different document types, language directions, edge cases and submission states — including unclear uploads, documents with sensitive content, and scenarios where translators attempted to skip the review step.
Design QA focused on ensuring the warning states, review prompts and approval flows worked as intended and that translators understood their role in the process before submitting.
Results
FirstPass helped reduce average document translation time from 37 minutes 54 seconds to 4 minutes 50 seconds, an improvement of approximately 87%, while keeping translators in control of review, correction and final approval.
The workflow helped refugees receive faster access to medical care, legal aid, housing support and case assistance — without removing the human judgment required for sensitive documents.
What this project reinforced
This project reinforced that AI product design is not about removing the human from the workflow. It is about designing the right relationship between automation and judgment.
FirstPass worked because the AI accelerated the first draft, while the translator still owned review, correction and final approval. The strongest design decision was not adding AI into the flow — it was designing the points where people could review, correct and approve before anything became final.
CRIBMD Healthcare Platform
Lead Product Designer · Healthcare ecosystem