An Introduction to Semaphore: Solving Problems with Semantics
summary
Semaphore claims it converts the stated 80% of enterprise data that is unstructured into searchable knowledge graphs and taxonomies, using three modules: taxonomy management, auto-classification (tagging), plus entity extraction and data harmonization exposed via APIs. Demos show mapping “moon buggy” to lunar roving vehicle, tagging an Apollo 11 article without the literal string, extracting market-cap figures from 10‑K PDFs, and enabling pivot‑table style BI on document corpora; it requires no document training set.
AI-written summary · about
chapters (10)
- 0:00Introduction to semantics
- 5:06Semantic challenges in business
- 8:59What is semantics in software
- 11:48Semaphore’s three services
- 18:50Breaking data silos semantically
- 21:43NASA mission demonstration
- 23:12Search as you type demo
- 27:46Apollo 11 document classification
- 30:33SEC 10K filings demo
- 35:55Q&A session
related talks
30mSemiont: A Graph Based, AI Native Wiki and Annotator
22mThe Capabilities of Semantic Search
31mComputers That “Get It”: Using LLMs to Give Software Semantic Understanding
22mStructuring the Unstructured : Empowering the AI Workflows
34mSemantic Introspection
01h 28mMaking Databases LLM-Ready Building Semantic Layers with Semantido
01h 28mFrom RAG to Riches Building Enterprise Search Systems with LLMs
32mDataEngBytes 2023 - PER-T1-11 - Mark Kitson