For my Campus Lost & Found Management System, I developed backend modules and implemented the claim-review workflow, from submission through review. I designed 15 tables and 32 APIs and documented their structure and indexes.
I also built matching recommendations using weighted scores for category, location, time, and keywords. In self-testing on 50 samples, the system achieved an approximately 70% hit rate.
On my Multimodal AI Agent Platform Based on Local Large Language Models project, I deployed Ollama and Open WebUI with Docker and configured local inference. I verified text, image-and-text, and speech-transcription scenarios, and documented common deployment issues.

