Practical Functional AI
Applied AI systems built from real-world problems.
AI Thought Lab is the applied AI workshop of Richard Coulson — a place for building, testing, and documenting practical AI systems across computer vision, automation, retrieval, and physical-world workflows.
Current Focus
- Image-based fabric identification
- OCR and label recognition
- Vector search and retrieval
- Factory and warehouse workflow tools
Featured Projects
Real systems, active prototypes, and practical AI workflows.
The center of AI Thought Lab is proof-of-work: projects that connect AI techniques to physical-world problems, messy inputs, and practical user workflows.
Active prototype
Computer Vision / Visual Search
Swatch ID
Image-based fabric identification using visual search, embedding models, and vector retrieval.
View case study →Working Prototype
Computer Vision / OCR
Bolt Finder
An OCR-first computer vision prototype for scanning warehouse-style rack images, extracting roll-number candidates, and highlighting matching fabric roll labels.
View case study →What This Lab Builds
AI tools that survive contact with real workflows.
The focus is not novelty AI. It is practical systems work: computer vision, retrieval, automation, OCR, local-first tools, and human-in-the-loop workflows built around real constraints.
Computer Vision
Image-based identification, visual matching, OCR, and label recognition.
Retrieval Systems
Vector search, embeddings, catalog lookup, and RAG-style workflows.
Workflow Tools
Human-centered AI utilities for factories, warehouses, samples, and process work.
Local Practical AI
Systems designed around useful results, privacy, speed, and real deployment constraints.
Proof Over Hype
Built from observation, tested against constraints, refined through use.
AI Thought Lab exists to document useful AI systems as they are built: what works, what breaks, what has to be simplified, and what becomes valuable once the tool meets the real world.
Lab Notes
Tech observations.
OCR-First Rack Scanning for Bolt Finder
How Bolt Finder moved from barcode-first assumptions to a working OCR-first rack scanning pipeline.
Computer VisionOCR vs Barcode Detection in Real Factory Photos
A short note on why label readability and image detail matter before choosing the recognition method.
Build Log
Current progress.
AI Thought Lab Reaches Pre-Deployment Milestone
The AI Thought Lab rebuild now has its core pages, project case studies, metadata, visuals, and launch structure ready for Cloudflare Pages deployment.
Bolt FinderBolt Finder Reaches OCR-First Proof of Concept
Bolt Finder moved from barcode-first assumptions to a functioning OCR-first rack scanning workflow.
Built by Richard Coulson
Applied AI from the perspective of a practical systems builder.
Richard brings together R&D experience, CAD and pattern workflow knowledge, factory-floor observation, automation thinking, and hands-on AI prototyping. AI Thought Lab is the public record of that work.