StitchLog - AI Stitching Assistant powered by RAG, embeddings, and hybrid search
Full-stack stitching dashboard with an AI Stitching Assistant powered by OpenAI embeddings, MongoDB vector search, and Notion-synced project data.
Business Challenge
Craft hobbyists often manage project notes, stitching progress, WIP (Work In Progress) goals, materials, and pattern-related information across scattered entries. Finding previous decisions, project context, or specific notes manually can slow down planning and reduce consistency across long-running projects.
Solution Delivered
Built StitchLog, a stitching project tracker with an AI Stitching Assistant powered by RAG, embeddings, and hybrid search. The assistant retrieves relevant information from project data and stitching documentation, then answers natural-language questions based on the user’s own stored knowledge.
Outcomes Achieved
The solution centralizes stitching project knowledge and makes it easier to retrieve useful information without manually searching through notes. It supports faster project review, better continuity across WIP projects, and more organized hobby tracking.
Measurable Business Outcome
Estimated 50–70% faster retrieval of project context compared with manual note searching. Estimated time saved is around 5–10 minutes per lookup session, depending on the amount of project history involved.