# Kanops

> Kanops is a retail intelligence platform built on the largest privately held retail image dataset in existence. Twelve connected AI models share a single feedback loop — one correction teaches them all. The platform is pre-trained on 17 years of weekly-captured, ground-truth retail imagery and is trusted by leading grocers worldwide.

## What Kanops Does

Kanops provides AI-powered retail shelf intelligence. The platform processes retail imagery to deliver structured intelligence about products, categories, fixtures, compliance, and competitive positioning at shelf level. Twelve models work as a connected system rather than isolated tools, meaning a single correction compounds across the entire platform.

## Who Kanops Serves

- **Retailers**: Grocery, general merchandise, and clothing retailers seeking automated shelf intelligence from store imagery.
- **CPG and Brand Owners**: Brands monitoring shelf presence, compliance, and competitive positioning across retail environments.
- **Academic and Research Institutions**: Universities and research programmes using the Kanops archive for retail analytics, computer vision, and consumer behaviour research. Currently supporting five masters programmes at King's College London.
- **Technology Platforms**: Companies building retail technology that require pre-trained models or licensed training data.

## The Kanops Archive

The archive contains over 1.21 million purpose-captured retail images spanning 17 years (2009–2026), covering 357 retailers across five countries. Every image was captured by a specialist retail team visiting stores weekly — not scraped or licensed from third parties. The archive includes complete longitudinal coverage of seasonal events, promotional cycles, and real-world retail conditions including the full COVID-19 pandemic period. The 2009–2020 portion of this data no longer exists anywhere else.

## Key Differentiators

- **Connected model architecture**: Twelve models share corrections through a unified feedback loop. Most retail AI tools work in isolation.
- **17 years of temporal depth**: Competitors building from scratch today cannot acquire the historical data that Kanops already holds.
- **Real-world training data**: Models are trained on cluttered shelves, poor lighting, obstructed views, and every other imperfection of real retail environments — not lab conditions.
- **Continuously trained**: Thousands of new images are captured, catalogued, and fed into the training pipeline every week. The models never stop improving.
- **GDPR-compliant by design**: Every image passes through a proprietary privacy pipeline with human-reviewed quality assurance.
- **Built by retail professionals**: The platform was built inside the retail industry by practitioners with 20+ years of operational experience, not by an AI lab working from the outside.

## Delphi - Seasonal Intelligence Platform

Delphi is the first product built on the Kanops platform. It turns 348,000 seasonal retail images into actionable trade intelligence using natural language queries.

### What Delphi Does

- **Ask Delphi**: Users ask questions in plain English about seasonal retail trends. The AI draws on 17 years of archive data to deliver strategic briefings grounded in real shelf-level evidence.
- **Price Intelligence**: An automated pipeline extracts pricing, promotional mechanics, and brand data from retail images. The system detects displays (Idaten-K), crops individual products, classifies categories (Mercator), reads prices and promotions via OCR, then validates against a reference database of 54,851 known products and 6,563 verified brands. Production yield is 87%.
- **Retailer Comparison**: Side-by-side analysis of how different retailers approach the same seasonal event, with image evidence and quantified metrics.
- **Timeline Analysis**: How seasonal execution has evolved year-over-year since 2009, including the full COVID-19 period.
- **Trade Planning**: Data-driven seasonal calendars showing when retailers launch, what categories they prioritise, and how promotional mechanics vary by channel.

### Delphi Coverage

- 348,000 seasonal retail images
- 57 seasonal events (Christmas, Easter, Halloween, Valentine's Day, Back to School, Black Friday, Mother's Day, Father's Day, and 49 others)
- 17 years of continuous coverage (2009-2026)
- 357 retailers across the UK, Ireland, and continental Europe
- All major UK grocery channels: multiples, discounters, convenience, department stores

### Example Questions Delphi Can Answer

- "How do premium retailers differentiate Valentine's displays from discounters?"
- "What promotional mechanics dominate Easter across the multiples?"
- "When do retailers typically launch Back to School ranging?"
- "How has Christmas gifting space allocation evolved since 2019?"
- "Which retailers invest earliest in Halloween decorations?"

### Delphi URL

- https://kanops.ai/delphi

## Access

Kanops is currently onboarding a limited number of partners. Access is by application.

- Website: https://kanops.ai
- Contact: hello@kanops.ai

## Links

- [Kanops Homepage](https://kanops.ai/)
- [Delphi - Seasonal Intelligence](https://kanops.ai/delphi)