Miros

Main Features

• Wordless Search: Finds any item without text entry needed.
• Inline Recommendations: Provides accurate recommendations for relevant products.
• Find by Image: Conducts visual product searches based on images.
• Discovery Bar: Semantic search locates items even when metadata is sparse or non-existent.
• Shop The Look & Style Collections: Offers discovery experiences for holistic outfit purchases and style collections.

Core Characteristics

• No "0 items found": Visual and semantic search algorithms find relevant items, completely eliminating the frustration of no results.
• No tagging: Semantic AI understands product photos and unstructured metadata via the Tagless Discovery system, requiring no manual tagging.
• Quick to start: Integration takes just 2 weeks, even for complex setups.
• Fast response: API response speeds are 50-300ms, with lean and optimized JS widgets that will not slow down website loading.

Target Users

Top retailers worldwide, especially e-commerce platforms in fashion, footwear, furniture, home decor, and stock images (e.g., Revolve, momox fashion, Ravensburger).

Core Advantages

• Deciphers both conscious and unconscious shopper behaviors: Accurately presents items that capture attention by analyzing scrolls, clicks, and searches.
• Less reliant on the quality of item metadata.
• Highly modular design: Can be used as core site search, product recommendations, or add-on widgets without replacing existing search providers.

Workflow

1. Data extraction: Rich information is extracted from product imagery to make it searchable.
2. Data enrichment: Visual search is combined with powerful semantic/NLP processing, blending traditional and AI search methods.
3. Easy integration: Integrated via GraphQL APIs and Web SDKs designed for modern headless e-commerce and traditional platforms.
4. Go Live: Visual AI algorithms track session behavior to improve matches based on user intent, ensuring real-time personalization.
5. Scale: Mastered scaling to ensure cost-effectiveness while delivering exceptional results.

Typical Results (Case Study Data)

• PERCENTIL: GMV +7%, AOV +4%, Retention +7%.
• Revolve: 4.5% GMV uplift among mobile shoppers.
• Carousell: 31% more "Buy Now" clicks.
• Ravensburger: 4.8x higher conversion, +14% avg. number of items bought.
• momox fashion: GMV +19%, AOV +11%, ARPU +10%.

FAQ

• Replace existing search: Not required. Miros is highly modular and can be used separately or together as core search, recommendations, or add-on widgets.
• Integration difficulty: The entire process typically takes just 2 weeks, requiring less than one engineering man-week on the client's end. Can be integrated via JS web widgets for quick setup or GraphQL API for greater flexibility.
• Impact on website speed: None at all. API response speeds are between 50-300ms, and JS widgets are lean and optimized.
• How to verify effectiveness: Before going live, a tailored instance trained on specific inventory is spun up to demonstrate the difference against the current solution.