Systems AI Business Management System Custom ERP Software Custom CRM Software AI ERP AI CRM Manufacturing ERP Software Jewellery Software Diamond Software Shopify Store Development App & Web App Development Replace Excel & Tally Yukti AI CRM Automation AI Automation Agency in Surat WhatsApp AI Chatbot WhatsApp Automation AI Agent Development Enterprise AI Agents Voice AI Agents n8n Automation Agency Workflow Automation GeM Tender Automation CRM Automation AI Consulting Industries Diamond & Jewellery Manufacturing Textile & Apparel Retail & E-commerce Travel & Tourism Real Estate Healthcare All industries Company Work / Case Studies Blog Free SEO Audit About Contact
Case study · AI image similarity search · Textile design · Surat

AI image similarity search for a Surat textile design house: find saree, suit and dupatta designs by photo or by words

How Yukti AI built textile design software that searches a design library by look. A fabric photo, or a description such as “pink floral border design”, returns the closest designs in the chosen category, each with its design code and a similarity score. Built in about 7 days on the client’s real files, and measured at every step.

A Surat textile design house makes print and embroidery-look designs for ladies’ ethnic wear, in a library of roughly 80,000 images across about 1,000 sets that could only be searched by folder and filename. Yukti AI profiled the real files, fixed the “colourway collage” problem with automatic main-panel cropping, tested OpenAI and Google Vertex AI embeddings head to head, chose Vertex AI, and delivered two visual-search apps plus an n8n search workflow. The system is designed to index the full library; the numbers here come from 97 recon files and a 401-design test catalogue.

Built in about 7 daysLive~80,000-design libraryRank 70 → rank 1
Executive summary

Quick facts

ClientA Surat textile design house (name withheld)
IndustryDigital print and embroidery-look designs for ladies’ ethnic wear: suit and kurti & dupatta sets, sarees, dupattas, lehenga, daman, all-over prints
LibraryRoughly 80,000 design images in about 1,000 sets, each design with a numeric code; the system is designed to index all of it
Measured on97 recon files in 2 sets; a 401-design test catalogue in 6 categories
Core challengeSearch only by folder and filename; files are colourway collages with inline watermarks; Google Lens cannot search a private catalogue
SolutionImage similarity search by photo or by words, inside one category, with design code and score on every result
Embeddings chosenGoogle Vertex AI multimodal (1408 dimensions), over gpt-4o-mini captions + text-embedding-3-small (1536)
Headline measurementWorst colourway-sheet case, design 13102: rank 70 of 97 raw, rank 1 after main-panel cropping
Build time & statusAbout 7 days; live, with two working search apps and an n8n workflow tested end to end
StackPython, Flask, Pillow, NumPy, Google Vertex AI, OpenAI, Pinecone, n8n
In short

What is image similarity search for a textile catalogue? Every design image becomes an embedding, a list of numbers that captures how it looks. A photo or a typed description becomes the same kind of vector, and the designs whose vectors sit closest are returned, whatever their filenames say.

What did Yukti AI build? Two live search apps that return the eight closest designs in a chosen category, one on Google Vertex AI multimodal embeddings (the pipeline chosen) and one on an OpenAI caption pipeline; an n8n search workflow over a Pinecone index; and a main-panel detector that isolates the motif on colourway sheets. Each pipeline change was scored as the rank of the true design among 97 samples.

The business

How does a Surat textile design house’s design catalogue work?

This design house produces digital print and embroidery-look designs for ladies’ ethnic wear. Every design carries a numeric code printed on the artwork, such as 13101 or 21875, alongside the brand watermark. Buyers choose by look; the library is filed by code. It is organised the way the trade works:

  • Sets — such as a viscose-look suits set with four colours per design, or an embroidery thread-stitching set of 30 designs.
  • Six categories — All Over, Saree, Kurti & dup (suit sets), Dupatta, daman and Lehengha.
  • Colourway sheets — one image with a large main panel plus 3–4 narrow strips of the same design in other colours.
  • Model mockups — for some designs, a model wearing the stitched kurti and dupatta.
  • Clean repeats — sets such as the embroidery set are full-bleed repeating tiles in one colourway.

The ask: when a customer sends a photo of a fabric or describes it, return the matching design code, plus similar designs in the same category.

The client’s rule: motif first, colour second

Every design already comes in several colourways, so “something pink” is not an answer. The design matters more than the colour, and that rule shaped every technical decision on this page.

The challenge

Why was an 80,000-design library impossible to search by look?

Because nothing in the library described what a design looks like. The only handles were folder names and codes inside filenames, and the files were made to show buyers, not to be read by software: collages, moving watermarks and uneven formats. The recon of real files found:

  • Collages, not designs. Most files are colourway grids and sometimes model mockups, so embedding a raw file captures “a four-panel collage”, not the motif.
  • Watermarks that move. The brand watermark and phone number are printed as text at varying positions, so no fixed crop removes them.
  • No two sets alike. Sizes from 1800×2250 to 4354×4000 px, 100, 150 or 300 DPI (sometimes within one set), aspect ratios from 0.8 to 1.1, and 0 of 97 files with a solid border to auto-crop on.
  • Uneven naming. Suffixes such as “Modeling”, “Suits 4 color” and “copy”, and mixed .jpg, .JPG, .jpeg and .jfif extensions.
  • Few lifestyle shots. About 7 of 67 codes in the suits sample had model photos, none in embroidery, so they cannot be the basis for matching.
  • Heavy files. About 1.48 MB each, or roughly 119 GB at 80,000 images; processing is limited by reading files more than by computation.
Colourway sheet and main panel
One image showing a design plus narrow strips of it in other colours; the main panel is the large section with the full motif, the part worth matching.
Motif, all-over repeat, pallu, buti
The drawn element of a print; a motif tiled across the fabric; the decorated end of a saree; a small scattered or corner motif.
Embedding, vector database, cosine similarity
Numbers an AI model produces to represent an image or sentence; a store that finds the nearest vectors fast; a score of how closely two vectors point the same way.
Rank of the true design
Where the correct design lands in the results list: the main accuracy measure in this build.
Before → after

How finding a design changed

Before

  • Designs findable only by folder name and filename
  • No search by what a design looks like
  • Google Lens could not reach a private catalogue
  • Colourway collages that confuse any whole-image match

After

  • Search by a fabric photo or a typed description
  • Eight closest designs in the category, each with code and score
  • Main panel of each colourway sheet isolated automatically
  • Design codes read from filenames; no catalogue image made public
The obvious question

Why not just use Google Lens?

The client raised it in a walkthrough, and it is the right question. Lens finds similar images on the public web, but it has no API for searching a private catalogue, so it cannot say which of a design house’s own unpublished codes a photo matches. The Google tool that can is Vertex AI multimodal embeddings, used for private vector search over your own files. That is the Google Lens alternative for a private catalogue, and it is what the final pipeline uses.

Requirement
Google Lens
Private visual search (built)
Searches your own unpublished designs
No, public web index only
Yes, your catalogue only
Search by a fabric photo
Against the web
Against your designs
Search by words (“pink floral border design”)
Web results
Text and images in one vector space
Returns your design code
No
On every result, with a score
Limit to one category (saree, dupatta, lehenga)
No
Category chosen first
API your own systems can call
None for private catalogues
n8n webhook: photo or text in
Step one: measure

What did the recon of real design files show?

Before any search code, Yukti AI profiled 97 real files from two sets, 67 suit colourway files and 30 embroidery files (143.9 MB), logging size, dimensions, aspect ratio, DPI, brightness and borders per file, plus a per-set profile of where the code and watermark sit. Aspect ratios clustered at 0.80, 0.91 and about 1.0. Then three sensible-sounding ideas were tested on real embeddings instead of being assumed:

HypothesisResult
H1: grayscale to make matching ignore colourSupported, with a caveat: motifs stayed distinct, but one design’s colourways collapse together
H2: index 30% crops so a partial photo can matchPartly: works for placement designs with landmarks, not dense all-over repeats; a crop must hold a full repeat
H3: find the set first, then the designLooked separable on 2 sets, too weak on real embeddings; so the user picks the category
The baseline

Why did the first AI model match the collage instead of the print?

Because a colourway sheet looks like a grid before it looks like a flower. A CLIP baseline (ViT-B-32) scored designs within a set closer to each other (0.816 and 0.841) than across sets (0.748): it was seeing each set’s template. With a model photo as the query, the plain design ranked anywhere from 5th to 51st of 60, and only 2 of 7 reached the top 10.

A caption pipeline came next: gpt-4o-mini describes the print, and text-embedding-3-small embeds the caption. With a 30% crop of a known design standing in for a customer’s photo, 3 of 4 queries on raw files put the true design in the top 5 of 97.

How accuracy was measured

A 30% crop plays the customer’s photo; the rank of the true design among all 97 is recorded; the pipeline is changed and the same queries re-run; scores are checked against an independent Python calculation. Every result here is a true design’s rank.

The fix

How main-panel cropping moved a design from rank 70 to rank 1

On the suit colourway sheets, a thin near-black or near-white vertical line separates the main panel from the colour strips. Yukti AI wrote a detector that scans a band of the image for that line in the 35–85% width range and crops everything to its left, so only the main panel is embedded. On the known failure case, design 13102, the true design moved from rank 70 of 97 to rank 1.

Scope was narrowed on purpose: full panel-splitting was estimated at 3–5 days of work against 2–3 for main-panel cropping, which fixed the measured failure directly, and the strips stay with their design as “also available in these colours”. Not every query improved: design 13401 dropped from rank 4 to 27 and is unresolved, and a controlled resize-only test is the next measurement.

70 → 1
Rank of design 13102 out of 97: raw colourway sheet versus main panel only
67 / 67
Suit colourway files where the divider was found, with no manual fallback
66.3%
Crop ratio shared by every file in the 13501–13508 batch; model images cropped at about 79–82%
Solution architecture

How the visual search system is put together

Category folders
Six categories; design code read from the digits in each filename
Prepare
Main panel cropped at the divider; images resized to 1024 px before captioning
Embed
Google Vertex AI, 1408 dimensions (earlier: caption + text-embedding-3-small, 1536)
Vector store
JSON cache for the apps; Pinecone with metadata for the n8n path
Query
Fabric photo or typed description, inside one category
Results
Top 8 designs with design code and similarity score
The head-to-head

OpenAI or Google Vertex AI: which embeddings found the right fabric designs?

On this client’s designs, Google Vertex AI. Both pipelines embedded the same 401-design test catalogue, and a harness listed the five nearest designs in the same category for six queries, one per category (Saree 6102, All Over 14260, Kurti & dup 16008, Dupatta 4530, daman 3170, Lehengha 576), checked by eye side by side.

OpenAI caption pipelineGoogle Vertex AI multimodal
Image to vectorgpt-4o-mini caption, then text-embedding-3-small (1536 dimensions)Image embedded directly (1408 dimensions)
Search by wordsText embedded like a captionSame model’s text encoder, same space as images
Rate limits200k tokens a minute hit after about 10 full-size imagesNone hit
All Over 14260 neighbours0.72–0.75, mixed designs0.84–0.89, its own 14255–14261 colourway block
Recorded API fee per imageAbout double the Vertex AI feeAbout half the OpenAI path’s fee
VerdictKept as its own app to compare or revertChosen

One spot check shows why this matters to a buyer: a magenta tropical-leaf dupatta (4530) returned the same illustration style in green (4851). That is design over colour, verified by eye. The fee comparison uses the cloud providers’ usage fees recorded during the build, not Yukti AI pricing.

What we built

The components of the visual search system

Beyond the recon inventory, main-panel detector and comparison harness above, these parts were built and run on real files.

Two catalogue builders

The OpenAI builder walks the six category folders, resizes each image to a maximum of 1024 px, has gpt-4o-mini describe it at temperature 0 (motif type, colours, density, layout style) and embeds the caption. The Vertex AI builder embeds each image directly with Google’s multimodalembedding@001 model. Both parse the design code from the filename digits, retry on rate-limit, quota and temporary service errors, and resume from a JSON cache; each produced 401 entries. Benefit: ingestion that survives interruptions.

Pinecone index and n8n visual-search workflow

A Pinecone index (1536 dimensions, cosine) holds the 97 recon embeddings with set, filename, design code, caption and a model-photo flag. In n8n, a webhook takes a JSON text query or a multipart image; a “Has Image?” branch runs Caption Image and Embed Caption for photos, or Embed Text for words, and both end at Query Pinecone. It was tested end to end live. Benefit: one endpoint a WhatsApp bot, website or ERP could call. More on n8n automation.

Two visual-search apps

Flask apps open on six category tiles with design counts, under a Gujarati-English prompt, “Sauthi pehla category select karo” (first, select a category). Each search page offers two options: upload a fabric photo, or describe it in words. Results are the top 8 designs in the category as thumbnails with design code and similarity score. The OpenAI app shows an “Interpreted as:” box with the caption; the Vertex AI app, badged “Google Vertex AI”, embeds photos directly and needs no captions for text search. They run side by side, so the team can switch or revert, and work on a phone. See custom web app development.

The workflow

How do you find similar saree, suit and dupatta designs?

Staff pick a category, then give the system a photo of the fabric or a few words describing it. The system turns that input into a vector, compares it with every design in the category and returns the eight closest designs with their codes and scores. Other systems can run the same fabric image search through the n8n webhook. AI steps are highlighted.

Search by photo

Customer sends a fabric photo
Staff pick the category
Photo uploaded
Embedded by Vertex AI
Cosine similarity across the category
Top 8 with design code and score
Staff act on the code

Search by words

Staff pick the category
Type “pink floral border design”
Text encoded into the image space
Top 8 matching designs

Search through the n8n webhook

Photo or text POSTed
“Has Image?” branch
Caption (gpt-4o-mini, temperature 0)
Embed caption or text
Query Pinecone
Nearest designs with metadata
Engineering that decides trust

How do you know the search results are right?

Compare the system’s scores with an independent calculation. Here, pipeline scores were checked against a separate Python reference, and the silent failures found in testing, a drifting caption and two n8n scoring bugs, were fixed during the build. A search that returns the wrong design with a confident score is worse than no search, so consistency, privacy and isolation from live systems were treated as requirements, not extras.

Temp 0

Same photo, same answer

Captions that varied between runs made one photo return different results. Found in testing and fixed with temperature 0 before customers saw it.

0.159

Silent bugs caught

Two n8n issues returned plausible but wrong scores with no error: 0.159 where 0.672 was expected. Only an independent Python check exposed them.

0.672

Exact match after the fix

A live text query then scored 0.672, matching the Python reference exactly.

Private

No public images

Photos travel as binary uploads, not hosted links, so no catalogue image had to be made public.

Isolated

Nothing live touched

New keys, index and workflow sit apart from the production bots, knowledge-base index and WhatsApp workflow in the same accounts.

1024 px

Built around limits

The caption model’s rate limit was hit after about 10 full-size images, so images are resized first.

Who uses it

What each person gets

Staff and sales

Pick a category, search by photo or description, and act on design codes with similarity scores, from a phone or a desk.

Customer

Sends a photo or description of the fabric they want; that is the query. A customer-facing channel such as WhatsApp is a separate step.

Catalogue admin

Adds folders of design images and rebuilds the embedding catalogue with the build scripts.

Textile-native data model

How should a textile design catalogue be labelled?

In the trade’s own words, not in generic image tags. Vectors find designs that look alike, but buyers still talk about pallus, butis and paisleys. Alongside the vectors, Yukti AI drafted a motif taxonomy for this design house, to be revised as more sets are seen, so labels use the vocabulary a buyer recognises:

FacetDraft values
LayoutAll-over repeat, border and pallu, placement panel, corner buti, garment-look mockup
Motif familyRealistic floral, linear-sketch floral, embroidery-texture floral, damask and lattice, geometric abstract, paisley, conversational
DensitySparse, medium, dense
BackgroundSolid, textured or grunge, gradient, dark ground, light ground
Colourway groupAn ID linking the colour variants of one design, derived at ingestion
Implementation process

How it was built in about 7 days

STEP 1

Recon of real files

97 files profiled for size, DPI, borders, codes and watermarks before any search code.

STEP 2

Hypotheses and baseline

Grayscale, crop indexing and set-first search tested; a CLIP baseline exposed the collage problem.

STEP 3

First pipeline, tested live

Caption embeddings, Pinecone and n8n end to end; silent scoring bugs and caption drift fixed.

STEP 4

Fix the collage problem

Main-panel detection: 67 of 67 dividers found; the worst case from rank 70 to rank 1.

STEP 5

Vertex AI pipeline

Direct image embeddings for the 401-design test catalogue, with text search in the same space.

STEP 6

Head-to-head and two apps

Six queries compared side by side; Vertex AI chosen; two search apps running.

Results

What was measured

Measured results only. No business outcomes, such as time saved, have been recorded yet, so none are claimed.

2nd of 97
Live photo query for design 21875 through n8n: score 0.933, just behind a visually similar sibling at 0.936
3 of 4
Partial-photo queries (30% crops) with the true design in the top 5 of 97, on raw files
0.84–0.89
Vertex AI neighbour scores for All Over 14260, all from its own colourway block
401 designs
Test catalogue: Kurti & dup 114, All Over 100, Saree 88, Dupatta 57, daman 31, Lehengha 11
Status

What is live today, and what comes next?

The system is live: two search apps run over the 401-design test catalogue, the n8n workflow runs over a 97-design Pinecone index, and Google Vertex AI is the pipeline chosen for the full library. This case study does not claim that the whole library of roughly 80,000 designs is indexed yet; the cards below separate what is live, what comes next and what is only possible.

Live today

Two search apps over the 401-design test catalogue, one on Google Vertex AI and one on OpenAI; the n8n workflow, tested end to end over a 97-design Pinecone index. Also built and measured: the main-panel detector and the Vertex AI catalogue builder.

Next steps

Index the rest of the roughly 80,000-design library with main-panel cropping at ingestion; run the resize-only test and resolve regressions such as 13401; revise the taxonomy.

Possible extensions

Yukti AI can connect the search to a WhatsApp bot or website, add filters such as fabric or stock status, and link results to an ERP. Options, not features already built.

Under the hood

Technology and security

Stack: Python · Flask · Pillow · NumPy for cosine similarity · Google Vertex AI multimodal embeddings · OpenAI gpt-4o-mini and text-embedding-3-small · Pinecone · n8n. An open-source CLIP model was used only for the baseline.

Security: no catalogue image was made public; the cloud project uses service-account authentication; API keys live in environment variables, not code.

Full technical detail
  • Vertex AI multimodalembedding@001 — 1408-dimension image and text embeddings in one space.
  • gpt-4o-mini + text-embedding-3-small — temperature-0 captions on 1024 px images; 1536-dimension embeddings.
  • Pinecone — dense serverless index, cosine, metadata per design.
  • n8n — Webhook, IF branch, OpenAI image analysis and HTTP Request nodes.
  • Flask, Pillow, NumPy, JSON caches — the apps, resizing, in-category similarity and resumable builds.
Commercial

What does AI image search for a textile business cost, and how long does it take?

It is a one-time custom build with a fixed quote, set after a free workflow audit of real sample files from your library. This one took about 7 days. Running costs are mainly cloud API usage and hosting, paid to the providers; embedding fees here were small per image. What moves the quote:

  • Library size — images and sets to ingest.
  • File condition — colourway sheets, mockups, watermarks, naming.
  • Query modes and categories — photo, words, or both.
  • Where search lives — staff app, website, WhatsApp bot or ERP.

Not sure AI search is the right first step? Start with AI consulting.

For other textile businesses

Can AI visual search work for your textile business in Surat?

Very likely, if your catalogue is large, visual and filed by code, whether it is a saree design catalogue, a suit and dupatta library or an all-over print archive, and it can be measured on your own files before you commit. The same approach fits print and embroidery design houses, saree, suit and dress-material traders who receive fabric photos, and dupatta, kurti and lehenga suppliers with designs in several colourways.

It is one part of what Yukti AI builds for the trade: see Textile & Apparel software, the self-serve B2B ordering system for a fabric-printing firm, the WhatsApp AI chatbot and AI agent development.

Yukti AI works from 220, Leonard Square, Yogichowk, Varachha, Surat, for businesses across Surat and India. Other Surat builds include a jewellery ERP for Lumera Fine Jewellery and a diamond trading platform for a Surat export house.

FAQ

Questions textile businesses ask about AI visual search

Yes, once the catalogue is indexed as image embeddings. For a Surat textile design house, Yukti AI built image similarity search that turns a fabric photo into a vector and returns the closest designs in the chosen category, with design codes and scores. The n8n search workflow accepts a photo upload, the same entry point a WhatsApp bot would call; connecting a live WhatsApp number is a separate step.
Google Lens searches Google's public web index and has no API for searching a private catalogue, so it cannot tell you which of your own design codes a photo matches. The Google tool that can is Vertex AI multimodal embeddings, which turns each of your designs into a vector that is searched privately. Yukti AI used it for this Surat design house without making any catalogue image public.
This client asked for motif first and colour second. Colourway strips stay with their design instead of becoming separate entries, and the pipeline isolates the main panel for matching. In a verified spot check with Google Vertex AI, a magenta tropical-leaf dupatta (4530) matched the same illustration style in green (4851), and All Over design 14260 returned its own colourway variants, 14255 to 14261.
Sometimes a close sibling ranks just above the exact design, so results show the top eight designs with codes and scores, and staff pick the exact code. In a live n8n photo test, design 21875 ranked 2nd of 97 with a score of 0.933, just behind a visually similar sibling at 0.936. For a buyer who wants the motif, that closeness helps: Google Vertex AI returned All Over design 14260’s own colourway variants as its nearest designs. The draft taxonomy adds a colourway group ID to link the variants of one design.
Not by hand. Yukti AI wrote a main-panel detector that finds the thin divider line between the main panel and the colour strips and crops automatically. It found the divider in 67 of 67 colourway files with no manual fallback, and the worst measured case, design 13102, moved from rank 70 of 97 to rank 1. Model mockups are rare, about 7 of 67 codes in the suits sample, so matching is built on the design files. Watermarks move around, so no fixed crop removes them; the ranks on this page were measured on the watermarked files.
Yes. Google Vertex AI encodes typed text into the same 1408-dimension space as the images, so a description such as 'pink floral border design' finds designs that have no caption or tags at all. In the OpenAI-based app, a photo is first described in words by gpt-4o-mini, and an 'Interpreted as:' box shows staff how it was read. Both apps take a photo or words.
Measure the rank of the true design. Here, a 30% crop of a known design stood in for a customer's photo, and the position of the true design among 97 sample designs was recorded before and after each change. On raw files, 3 of 4 queries put it in the top 5; after main-panel cropping the worst case moved from rank 70 to rank 1. Scores were checked against an independent Python calculation.
On this client's designs, Google Vertex AI. Yukti AI ran both on the same 401-design test catalogue and compared the five nearest designs for six queries side by side. For All Over design 14260, Vertex neighbours scored 0.84 to 0.89 and matched its own colourway block, while OpenAI caption embeddings scored 0.72 to 0.75 and were mixed. Vertex also cost less per image and hit no rate limits.
The build is a one-time project with a fixed quote after a free workflow audit, depending on library size, file condition, categories and channels. Running cost is mostly cloud API fees paid to the provider, and embedding fees are small per image. In this build, the recorded Google Vertex AI fee per image was about half that of the OpenAI caption path, with no rate limits hit. For a very large library, reading heavy image files takes more time than the AI does.
No. In the n8n workflow Yukti AI built, a query photo travels as a binary file upload rather than a hosted link, and catalogue images are embedded straight from the client's own files. The vector index keeps only vectors with metadata such as set, file name, design code and caption. No design had to be published on the web for the search to work.
Yes, by design. The app opens on six category tiles, All Over, Saree, Kurti & dup, Dupatta, daman and Lehengha, each with its design count, and search runs only inside the chosen one. Testing showed that separating sets automatically from the image alone was too weak, so the person searching picks the category. Results are the top eight designs, each with code and score.
This one was built in about 7 days: recon of real files, hypothesis tests, a CLIP baseline, an OpenAI caption pipeline with a Pinecone index and n8n workflow, main-panel cropping, a Vertex AI pipeline, a head-to-head comparison and two search apps. Indexing a very large library afterwards is mostly processing time, limited by reading large image files more than by the AI.
Yes, and isolation was a rule of this build. The search runs as its own n8n workflow with its own webhook, keys and vector index, apart from the production bots and WhatsApp workflow already running in the same accounts. The webhook accepts a photo upload or a JSON text query, the shape a WhatsApp bot or website would call, so connecting one is a deliberate later step.
Case study at a glance

Summary

Knowledge summary
  • Yukti AI, a software company in Varachha, Surat, built AI image similarity search for a Surat textile design house in about 7 days, to find designs by a fabric photo or a typed description.
  • The Surat design house’s library holds roughly 80,000 designs in about 1,000 sets; Yukti AI designed the pipeline to index all of it and measured it on 97 recon files and a 401-design test catalogue.
  • Yukti AI’s main-panel detector for colourway sheets found the divider line in 67 of 67 files and moved the worst case, design 13102, from rank 70 of 97 to rank 1.
  • In Yukti AI’s head-to-head test on the client’s own designs, Google Vertex AI multimodal embeddings beat an OpenAI caption pipeline on motif matching and cost, and were chosen.
  • Yukti AI’s textile visual search apps return the top eight designs inside a chosen category, such as saree or dupatta, each with design code and score.
  • Yukti AI’s n8n search workflow for the Surat design house found design 21875 at rank 2 of 97 in a live photo test and matched a Python reference score of 0.672 exactly in a live text test.

Still finding designs
by folder and filename?

Tell us how your design library is organised — sets, categories, colourway sheets and codes. In a free workflow audit we’ll look at real sample files and show you how image similarity search would work on your own catalogue.

Free audit · No commitment required · Surat, Gujarat