Notion Adds Speaker ID to Meeting Notes

Notion Adds Speaker ID to Meeting Notes

Notion quietly rolled out speaker identification for its AI Meeting Notes feature, addressing what the company calls a "top user request" [4]. The update works for 1:1 meetings and some video conferences, marking Notion's first step into diarization research.

CEO Ivan Zhao's repost of the announcement suggests this is just the beginning of Notion's push into meeting intelligence [5]. The feature aims to improve transcription accuracy by clearly attributing who said what during conversations.

Open-Source RAG Framework Ditches Vector Databases

VectifyAI released PageIndex, an open-source RAG framework that achieves 98.7% accuracy on complex financial documents without using vector databases [6][7]. Instead of traditional embeddings and chunking, PageIndex builds hierarchical tree structures that let LLMs navigate documents more intelligently.

The framework particularly excels with structured documents like SEC filings, where traditional vector RAG typically manages only 30-50% accuracy [8]. For meeting transcription companies dealing with complex, multi-topic conversations, this tree-based approach could offer a more nuanced way to organize and retrieve conversational knowledge.

EU AI Act Compliance Deadline Approaches

European companies have until August 2, 2026 to comply with the EU AI Act's requirements for high-risk AI systems [9][10]. This includes AI used for employment decisions, credit scoring, and fraud detection — areas that increasingly overlap with meeting intelligence platforms analyzing workplace conversations.

The penalties are substantial: up to €35 million or 7% of global annual turnover, whichever is higher [11]. Companies must demonstrate explainability, risk management, and human oversight for covered AI systems.

What This Means For Your Meetings

The meeting intelligence space is rapidly maturing beyond simple transcription. Otter.ai's pivot to enterprise knowledge graphs and Notion's speaker identification push reflect a broader shift toward treating conversations as interconnected knowledge assets rather than isolated events. The question isn't just "what was discussed?" but "how does this connect to everything else we know?"

PageIndex's success with tree-based document navigation hints at more sophisticated ways to organize conversational data. Traditional vector search works well for finding similar content, but meetings often contain multiple topics, decisions, and context shifts that might benefit from hierarchical organization. Meanwhile, the EU AI Act's August 2026 deadline means European companies need to start thinking seriously about explainability and oversight for any AI that touches employment or business decisions.

Key takeaway: Meeting intelligence is evolving from a productivity tool into enterprise knowledge infrastructure, but with that power comes new compliance responsibilities.

Sources

1. https://otter.ai/blog/otter-ai-evolves-from-ai-notetaker-to-create-100b-enterprise-conversational-knowledge-engine-market
2. https://finance.yahoo.com/sectors/technology/articles/otter-ai-evolves-ai-notetaker-120000195.html
3. https://gilbane.com/2026/04/otter-ai-launches-enterprise-conversational-knowledge-engine
4. https://x.com/NotionHQ/status/2051699415463956918
5. https://www.notion.com/product/ai-meeting-notes
6. https://github.com/VectifyAI/PageIndex
7. https://pub.towardsai.net/pageindex-the-rag-framework-that-threw-out-vector-databases-and-still-hit-98-7-accuracy-d194e0549478
8. https://venturebeat.com/infrastructure/this-tree-search-framework-hits-98-7-on-documents-where-vector-search-fails
9. https://ai-act-service-desk.ec.europa.eu/en/ai-act/timeline/timeline-implementation-eu-ai-act
10. https://artificialintelligenceact.eu/implementation-timeline
11. https://www.compliquest.com/en/blog/what-is-eu-ai-act-requirements-2026