Dear Proton Development Team,

First, I want to express my sincere appreciation for the incredible work you do. As a long-time Proton Unlimited user, I deeply value your commitment to privacy, security, and the "privacy-first" philosophy that sets you apart in the tech landscape. The fact that Lumo is built on zero-access encryption and respects user data is exactly why I chose to adopt it.

I am writing today to share a piece of feedback regarding a potential feature enhancement that I believe would significantly boost Lumo's utility for daily life: Optical Character Recognition (OCR) for image uploads.

Currently, while the file upload feature works seamlessly on my account, I am unable to extract text from images (such as photos of receipts, fuel pump displays, or invoices). For many users, including myself, this is a missed opportunity for powerful automation.

Why this matters: Imagine a workflow where I could simply snap a photo of a fuel pump display or a grocery receipt, upload it to Lumo, and have the AI instantly:

Extract the date, total amount, currency, and quantity.
Automatically format this data into a CSV or spreadsheet row.
Categorize the expense and update my budget or fuel tracking logs (like my existing fuelings.csv).
This would turn Lumo from a conversational assistant into a true productivity partner for personal finance and logistics. It would save hours of manual data entry and reduce the friction of tracking daily expenses.

A Note on Privacy: I understand that Proton prioritizes user privacy above all else. I am confident that this feature could be implemented in a way that aligns perfectly with your values:

Local Processing: Ideally, the OCR processing could happen locally on the device (if supported) or in a zero-knowledge environment where the image is processed and the text is extracted without the raw image data ever being stored or logged on your servers.
Explicit Consent: The feature could be opt-in, requiring explicit user permission for each image analyzed, ensuring full user control.
I believe that adding OCR capabilities would not compromise the security model you have built but would instead demonstrate how privacy and powerful utility can coexist. It would make Lumo an indispensable tool for anyone trying to manage their life efficiently without sacrificing their data sovereignty.

Thank you for listening to this suggestion and for continuing to build tools that empower users. I look forward to seeing how Lumo evolves and hope this feature finds its way into a future update.

Warm regards,

Heiko