## TL;DR
AI is the newest member of every team -- transcribing meetings, extracting action items, surfacing relevant documents, and connecting colleagues who should be collaborating. From Microsoft Teams Copilot to Notion AI, AI-augmented collaboration reduces meeting overhead by 10+ hours/month while improving knowledge sharing.

## Core Explanation
AI collaboration stack: (1) Meeting intelligence -- real-time transcription (Whisper, Azure Speech), speaker diarization (who said what), summarization (extractive + abstractive), action item extraction (NLP: "I will send the report by Friday" -> Action: Send report, Deadline: Friday, Owner: Speaker), and sentiment analysis (meeting tone); (2) Knowledge management -- AI indexes organizational documents (Google Drive, SharePoint, Notion). RAG-based Q&A: "What was the Q2 marketing strategy?" retrieves and synthesizes; (3) Expertise location -- NLP profiles employee skills from documents, emails, code commits, and meeting transcripts. Query: "Who knows about Kubernetes deployment?" returns ranked experts; (4) Collaboration analytics -- AI analyzes meeting calendars, email volume, and chat patterns to detect overload, silos, and burnout risk.

## Detailed Analysis
Meeting intelligence: Whisper (OpenAI) provides state-of-the-art transcription. Speaker diarization (Pyannote) segments audio by speaker. LLM summarization: prompt with full transcript -> extractive (key points) + abstractive (concise summary). Factuality challenge: meeting summaries must not hallucinate decisions. Solution: extractive verification (every summary claim linked to transcript quote). Microsoft Teams Copilot: integrated across Outlook (draft emails, summarize threads), Teams (meeting recap, action items), and Office (document drafting). Google Duet AI: meeting notes, real-time translated captions (70+ languages), and document Q&A. Notion AI: Q&A over team wiki, auto-fill database properties, and writing assistance. Organizational network analysis (ONA): graph where nodes are employees and edges represent collaboration (co-authorship, email, meetings). GNNs predict information flow bottlenecks and identify key connectors. Key challenge: privacy -- AI processing of all communications raises surveillance concerns. Opt-in, transparency, and data minimization are essential.