Lot Lingo: AI-Powered Auction Workflow Platform
The Challenge
Auction houses face a recurring administrative burden every time they prepare an auction: researching items, writing titles and descriptions, categorizing lots, assigning taxability and shippability, organizing photos, and preparing platform-specific exports. For high-volume auction businesses, this work creates a major bottleneck that limits how many auctions can be run and which lower-value items are economically worth listing.
Users
Auction house staff (cataloguers, photographers, owners) at 15 paying B2B customers, plus their downstream bidders.
Workflow
Photo capture → AI-generated titles, descriptions, categories, taxability, and shippability → human review and approval → platform-specific export to K-BID or HiBid.
Constraints
Auction listings affect buyer expectations, seller relationships, taxability, and shipping obligations, so AI output must be reviewable and correctable. Workflows must work on mobile in the field, support voice input, and integrate with the platforms customers already use.
AI Capabilities Implemented
Image Analysis
Computer vision for photo-driven lot creation, including categorization and shippability assessment from images.
Document Transcription
In-app image editing and background removal for consistent listing presentation.
Voice Transcription
Voice notes captured while photographing lots; voice-driven edits and information capture.
Additional AI Features
- AI-generated lot titles and descriptions tailored to auction conventions
- Auction-level title and description generation from full lot context
- Automatic categorization aligned to K-BID and HiBid taxonomies
- Taxability assignment with human-reviewable defaults
- Shippability assignment that flags oversized or impractical items
- Collaborative AI assistant that can revise lots and add information conversationally
- Human-in-the-loop review and approval at every stage
Results
Technical Architecture
Frontend
Web and mobile-first capture workflows with collaborative editing
Backend
AI orchestration layer with workflow state, voice, and image pipelines
Database
Operational database for auctions, lots, customers, and review state
Storage
Cloud object storage for photos and processed image artifacts
Integrations
- • K-BID auction platform export workflows
- • HiBid auction platform export workflows
- • Multiple AI vendors (vision, text generation, voice)
- • Image editing and background-removal services