# 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. Over 1.7 million cross-model learning entries ensure every model benefits from corrections made anywhere in the system.

## 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 1.22 million purpose-captured retail images spanning 17 years (2009-2026), covering 360 retailers predominantly in the UK, with coverage in Ireland, the US, Germany, and the Netherlands. 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 with over 1.7 million cross-learning entries. 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 364,000 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, crops individual products, classifies categories across 20 retail categories, reads prices and promotions via OCR, then validates against a database of over 54,000 known products and hundreds of verified brands.
- **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

- 364,000 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)
- 360 retailers predominantly in the UK, with coverage in Ireland, the US, Germany, and the Netherlands
- 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)
- [Seasonal Events Archive](https://kanops.ai/seasonal-events)
- [Full Image Archive](https://kanops.ai/full-archive)