AI‑powered visual product discovery API for design‑forward eCommerce

Add multimodal search to your storefront: type a sentence, paste an image URL, or upload a photo. Our API extracts meaning and structured filters—then returns clean, de‑duplicated results that feel like it’s reading your mind.

Chat Image Search
curl -X POST https://api.designview.ai/api/v1/chat/search \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{
    "query": "matte black ceiling fan with light from Minka Aire, modern, 52in",
    "image_url": null,
    "options": { "extract_nlp_filters": true, "deduplicate_visuals": true }
  }'

Replace dozens of product filters with one conversation

Filter‑based Search

Modern Indoor and Outdoor Ceiling Fans — 2,474 Results
Finish: Black Style: Modern Size: 52"
Category
Ceiling Fans
Mount type
With light kit
Brand
Finish
Blade span
Approximately 52"
Price
Motor type

Lots of steps, toggles, and decisions—clunky user experience

Semantic Chat Search

API-based Concierge — 50 Curated Results
Looking for a matte black ceiling fan with a light from Minka, modern, about 52 inches wide
Great — you want a matte black, modern Minka Aire ceiling fan with a light around 52", perfect for a medium‑sized room like a bedroom or living area; I’ll look for sleek, contemporary options in that finish and size.
But only with DC motors
Got it — you want fixtures that use DC motors. I’ll look for ceiling fans and other motorized fixtures with DC motors, focusing on the size, style, and finish you prefer; if you tell me the room, I’ll suggest which fan types work best there.

We extract structured filters automatically—no manual toggles.

We use computer vision models — The future of product image search

Modern ceiling fan with integrated LED light

Input Text or Image

Upload a product photo or describe what you're looking for. Our AI converts both images and natural language into semantic embeddings.

Product discovery process visualization

Generate Embeddings

Your input is transformed into high-dimensional vectors that capture visual features, style, and semantic meaning in a unified embedding space.

Visual search with multiple attributes

Find Nearest Matches

Using a machine learning model, we instantly find products with similar embeddings, returning visually and semantically related items with extracted attributes.

DesignView API — Feature highlights

AI-powered product discovery

  • Multimodal search from text, image URL, or uploaded image
  • Smart query optimization for better visual matching
  • Automatic NLP filter extraction (dimensions, types, mount, light source, safety ratings, brand include/exclude, exact product IDs)

Cleaner, more relevant results

  • Visual de-duplication returns unique visual groups
  • Sensible defaults exclude low-intent categories unless requested
  • Special logic for common intents (e.g., fans vs. controls, wet rated rooms)
  • Computer vision for elements missing from your data (e.g., DC motors)

Guided shopping experience

  • Customizable, friendly assistant replies that acknowledge user criteria and offer more style guidance
  • Image analysis to textually describe style, materials, and form factor

B2B brand access controls

  • Per-customer manufacturer allowlists enforced at the API layer
  • Prevents showing products from unapproved brands in a specific e-commerce site

Administrative management endpoints

  • Replace or partially update customer permissions
  • Automatic backups for customer permission updates; instant cache invalidation
  • Secured via admin API keys

Reliability, performance, and security

  • Fast vector search backend for scale
  • Health and status endpoints for monitoring
  • Global rate limiting, CORS, and no-cache headers
  • Robust error handling and structured JSON responses

Ready to build?

Integrate DesignView AI into your product search experience with a few lines of code. We handle your product image embeddings and vector search so you can focus on your product catalog and user experience.

Get API access for your environment, or ask us to see a demo.