Pattern: Interior Shopping List From Product Photos

Status: draft / demo-candidate
Module: Automation Atlas
Category: Operations / Documents / Field Work
Target industry: interior designers, architects, home furnishing coordinators
Source: ../intake/2026-06-07-ankieta-marta-automation-shopping-list.txt
Related branch: ../industries/interior-design-procurement.md
Last reviewed: 2026-06-07

Business Problem

Interior and furnishing projects require many product decisions from physical stores. A designer or coordinator collects photos, prices, tags, categories and notes, then prepares a readable shopping list for a client.

Manual preparation is repetitive and easy to break because product data comes from photos, shop tags, changing prices, human notes and project-specific room naming.

Trigger

The automation starts when a user adds a new project and uploads product photos or fills a product intake form after a store visit.

Inputs

• project name,
• investor/client email,
• investment address,
• product photo,
• shop name or visible shop tag,
• product category,
• product name or tag text,
• brand,
• price,
• unit price,
• quantity,
• barcode,
• product link,
• room assignment,
• color/pattern,
• free-form note,
• photo date or store visit date.

Outputs

• online client-facing shopping list,
• grouped view by room,
• grouped view by shop,
• total budget summary,
• placeholders for missing prices or links,
• product table for review,
• optional export to spreadsheet or PDF later.

Systems Involved

• mobile photo capture or upload,
• OCR / vision extraction layer,
• product data schema,
• human review UI,
• online list / client share link,
• optional spreadsheet,
• optional n8n workflow for notifications and follow-up.

AI Role

AI can:

• read product tags from photos,
• detect shop when visible on tag,
• extract approximate and unit prices,
• classify product type,
• propose product name and brand,
• identify whether a tile is floor, wall or both,
• create placeholder records when data is incomplete,
• generate clean client-facing descriptions.

Human Review Point

A human must approve:

• price correctness,
• final quantity,
• room assignment,
• product category,
• product link,
• budget summary,
• client-facing final list.

Implementation Complexity

Medium.

The first version can be simple if it avoids full product search and focuses on structured capture, review and presentation. Complexity grows if automatic product matching, barcode lookup, surface calculation and stock validation are added.

Risk Level

Medium.

Risks

• OCR may misread price tags.
• Product prices can change.
• Product links may not be easy to find automatically.
• Photos may not show all needed data.
• Room naming is project-specific.
• Quantities and surface calculations require human responsibility.
• Client-facing list must be visually clear and trustworthy.

Example Use Cases

• Bathroom furnishing list with tiles, fixtures, grout and accessories.
• Whole-home furnishing list with panels, trim, appliances and room-specific products.
• Store visit capture for Leroy Merlin, Castorama and Komfort.
• Premium interior design client list shared as an online link.

Reusable Assets

• product intake questionnaire,
• shopping-list item schema,
• OCR/vision prompt,
• review table UI,
• client-facing online list template,
• n8n reminder flow,
• RS360 demo scenario.

MVP Shape

1. User creates project and enters investor email plus address.
2. User uploads product photos and optionally types notes.
3. AI extracts visible fields and creates draft rows.
4. User corrects rows in a review table.
5. System generates online list grouped by room and shop.
6. Missing prices or links stay as clear placeholders.

Source Observations

Important discoveries from Marta's questionnaire:

• Both regular price and unit price may be needed.
• Price can be treated as approximate at first.
• Product categories must be broad enough for whole-home furnishing.
• Floor vs wall tile distinction is mandatory.
• A product may belong to floor and wall at the same time.
• The first list should support both room and shop grouping.
• The first client output should be an online link.
• Text areas in forms should be visibly larger, not tiny one-line fields.
• The intake should collect project name, investor email and investment address at the start.

Next Actions

• Compare with formularz_beta_lista_zakupow_v1.html when imported.
• Define shopping_list_item schema.
• Decide whether this becomes an RS360 demo.
• Create a simple UI mockup after the Planner UI rebuild phase.