Public portfolio project · Portfolio reviewed October 2026
AI chess scoresheet parser
A portfolio application that turns photographed or scanned handwritten chess scoresheets into structured moves, validates them with chess rules, and supports replay on an interactive board.
The work
A handwritten scoresheet is an unstructured input; a replayable game needs structured moves in a valid sequence. This project combines extraction with domain-specific validation and an interface for inspecting the result.
- Scanned or photographed handwritten scoresheets are the stated input.
- The portfolio describes OCR and move extraction followed by validation against chess rules.
- An interactive board lets the user replay the extracted game. The portfolio lists Next.js and FastAPI among the project technologies.
Public record
Published as an Upwork portfolio project with a linked application. It is not presented as a paid client case or an independently measured OCR benchmark.
Source: AI Chess Scoresheet Parser - OCR, Move Extraction. Public Upwork profile reviewed October 27, 2026.
View the original Upwork record ↗Open the portfolio application ↗External application; use a non-sensitive sample if you choose to try it.
Planning similar work?
The same design question applies to document intake: what structure should be extracted, which rules can validate it, and how should a person handle an uncertain result? The document-intake example below demonstrates those decisions with synthetic data.
Try the synthetic document-intake example →