Services
Turn research and development assets into systems people can actually use.
I support AI search, RAG preparation, agriculture and measurement workflows, legacy C/Java repair, image-recognition environments, and practical web dashboards. The work can start from one document, one CSV file, one screenshot, or one old source-code folder.
AI search and RAG preparation
Organize documents, research notes, manuals, and CSV files before introducing LLM search.
- Classify internal documents and research material into searchable units
- Prepare data structure, keywords, synonyms, and chunking rules for RAG
- Create evaluation notes, improvement plans, and operating procedures
Agriculture and measurement DX
Build practical tools for field measurement, barcode workflows, Android devices, and CSV operation records.
- Android apps and Bluetooth/HID/SPP/BLE device integration
- Project, template, phenotype, and CSV management
- Progress, duplication, remaining time, and histogram visualization
Legacy C/Java asset repair
Read old source code, restore build environments, and separate what should be repaired, migrated, reused, or retired.
- C/C++ memory issues, type expansion, and Visual Studio migration
- Java/SWT/Tomcat investigation and migration planning
- Source cleanup, runbook, and verification report
Image recognition and federated learning
Prepare small but reproducible AI environments for YOLO, Flower, Docker, and test data.
- YOLO training/inference environment and test dataset preparation
- Flower server/client structure for federated-learning experiments
- Dockerfile, compose.yaml, and execution procedure documents