01
the user adds 2–5 photos of one item to the working folder
Product case study
PstrykOpis: from item photos to a ready Vinted listing
PstrykOpis connects analysis of 2–5 photos, listing-data preparation, and a Chrome extension in one controlled workflow. The tool helps fill the Vinted form, while the final publishing decision stays with the user.
Key implementation elements
analysis of 2 to 5 photos of one item with the Gemini API
suggested title, description, category, attributes, and price for review
local analysis history and separate folders for workflow stages
Chrome side panel with listing search, list view, and field details
inserting the full field set or selected values into the Vinted form
manual review and publishing controlled by the user
Problem before implementation
Preparing each listing manually repeats the same set of steps
For every item, the seller has to review the photos, identify the category and product attributes, write a title and description, estimate a price, and transfer the data into the form. Without one workflow, information easily drifts between photos, notes, and the browser.
repeatedly describing an item from several photos
manual preparation of the title, description, category, and listing attributes
copying many fields between an application and the Vinted form
no local history for returning to earlier analyses
What we built
A Windows app and Chrome side panel working on one data set
We built a local tool that reads photos from a working folder, sends them to Gemini for analysis, and saves the prepared listing in JSON history. The Chrome extension reads that data through a local API and helps insert it into the Vinted form without taking over final publication.
Process after implementation
How the process works after implementation
From input data to a cleaner outcome. Below is a shortened view of the process after implementation.
02
the application validates the photo set and sends it to Gemini for analysis
03
a suggested title, description, category, attributes, and price are prepared
04
the result is stored locally and appears in the Chrome side panel
05
the user inserts selected fields into the form, reviews them, and publishes the listing manually
Before / after
How the process changed
The table shows the main differences between manual work and the process after implementation.
| Before implementation | After implementation |
|---|---|
| description and attributes prepared separately for every item | one analysis result groups listing fields into a consistent set |
| photos, notes, and the form operate as separate steps | working folder, history, and browser extension form one workflow |
| many fields copied manually one by one | the full set or selected fields can be inserted into the form |
| automation could take over too many decisions | the user retains control over corrections and publication |
Process outcome
less repetitive work between source photos and an empty listing form
one consistent set of suggested listing data for review
local history for returning to earlier results
automation support without giving up control over publication
Technologies
Windows desktop applicationGemini APIimage analysisChrome Extension Manifest V3local APIJSON history
What can be implemented in a similar way
What can be implemented in a similar way
These are examples of processes that can be organized with a similar approach: start from one concrete problem and a clear data flow.
product-photo analysis and catalog-data preparation
listing-description assistants with human approval
local tools connecting AI with a browser extension
photo-to-data workflows ending in a controlled form
Related services
This type of implementation can be connected with MorenaTech's core areas
If a similar process still runs manually or is scattered across files, it can be connected with automation services, Google Workspace, or further process development.
Final CTA
Want to connect photo analysis with a controlled workflow?
If your team manually transfers information from photos into forms or catalogs, a similar process can combine AI, local history, and a human who approves the result.