Online Meeting Effectiveness Index (OMEI)

2026-06-09 · Source: Microsoft Foundry Blog articles · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Advanced, long

Summary

The Online Meeting Effectiveness Index (OMEI) is an LLM-based framework developed by Santa Clara University in partnership with Microsoft to score and evaluate virtual meetings. Utilizing Azure AI, OMEI assesses meetings across five dimensions: Participation, Engagement, Structure, Sentiment, and Tech Quality, each with three sub-metrics. The system processes public video recordings via Azure Video Indexer for transcription and metadata, then uses GPT-5.4-mini within Microsoft Foundry for LLM-based analysis and scoring. An analysis of 68 real meetings revealed an average OMEI score of 86/100, with Tech Quality scoring highest at 92.3 and Engagement lowest at 77.3. Initial LLM scores were calibrated against human reviews, reducing the average from 87.1 to 72.5, aligning more closely with human judgment.

Key takeaway

For meeting organizers aiming to improve virtual collaboration, OMEI demonstrates how AI can provide objective, actionable feedback. You should prioritize clear agendas, defined objectives, and explicit action items to boost engagement and structure, which are common weaknesses. AI Engineers can integrate similar LLM-based frameworks into existing platforms like Microsoft Teams, utilizing Azure services and calibrated prompts to transform unstructured meeting data into measurable effectiveness insights for continuous improvement.

Key insights

LLMs can effectively score virtual meeting effectiveness across key dimensions, offering actionable improvement insights.

Principles

• Psychological safety predicts meeting success (OR=3.3).
• Speaking 10% of meeting time increases inclusiveness 4x.
• Unclear goals are major productivity obstacles.

Method

OMEI's pipeline ingests video, extracts transcripts and metadata via Azure Video Indexer, then uses GPT-5.4-mini in Microsoft Foundry with calibrated prompts for LLM-based scoring across five dimensions, outputting a score and recommendations.

In practice

• Define clear agendas and objectives.
• Build intentional interaction points in presentations.
• Summarize decisions and assign action items.

Topics

• Online Meeting Effectiveness Index
• Large Language Models
• Azure AI
• Virtual Meeting Analytics
• Meeting Effectiveness Scoring
• Organizational Productivity

Best for: Executive, Research Scientist, AI Product Manager, AI Scientist, AI Engineer, Director of AI/ML

Related on AIssential

• 🧮 OpenAI started rolling out sponsored ads inside ChatGPT for users in the U.S. on the Free and Go tiers. — AI integration is reshaping business operations, job markets, and model evaluation, while also introducing new monetization and workflow dynamics.
• How Intelligent Machines Are Redefining Success — Collaboration with AI is crucial for success in the "AI-First era" across business, creativity, and research.
• When large language models are reliable for judging empathic communication — LLMs reliably judge empathic communication, matching experts and outperforming crowdworkers when properly benchmarked.
• Design a Reliable LLM-Integrated Interface for Mortality Forecasting — LLMs can enhance complex analytical tools' accessibility without sacrificing validity or transparency.
• Does Theory of Mind Improvement Really Benefit Human-AI Interactions? Empirical Findings from Interactive Evaluations — Static ToM benchmarks do not predict LLM performance in dynamic human-AI interactions.
• Building a Self-Improving AI Support Agent with Langfuse — Langfuse enables robust LLM production systems through comprehensive tracing, metrics, and user feedback mechanisms.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Microsoft Foundry Blog articles.