# MicroStrategy Acquires 18,300 BTC and achieves Yield of 4.4% QTD and 17.0% YTD Source: https://www.microstrategy.com/press/microstrategy-acquires-18300-btc-achieves-btc-yield-of-4-qtd-17-ytd-now-holds-244800-btc Kiwi News: https://news.kiwistand.com/stories/MicroStrategy-Acquires-18300-BTC-and-achieves-Yield-of-44percent-QTD-and-170percent-YTD?index=0x66e430cb983f329ed041aa9ec200e3d4566df09ee72ea7e4e18cab7062e311dee90fb935 Submitted by tim on 2024-09-13 UTC, 1 upvote, 0 comments ## Summary Strategy Mosaic, a universal semantic layer from Strategy, is presented as a way to make analytics and AI more reliable, governed, and cost-efficient by defining governed metrics and access policies once across BI, applications, and AI. In the company's benchmark, Mosaic answered with 100% accuracy, against 88.2% when querying raw data, and it is said to use up to 53% fewer tokens. Pfizer is named as running 15,000 users across 27 markets on Mosaic. The text does not mention the bitcoin acquisition or yield figures in the title. ## Article Why Mosaic Strategy software helps enterprises make analytics and AI more reliable, governed, and cost-efficient by giving people, applications, and AI agents the same trusted business context. At the core is Strategy Mosaic, a universal semantic layer that defines governed metrics, relationships, calculations, and access policies once and applies them consistently across BI, applications, data products, and AI. Mosaic works with the clouds, databases, and BI tools you already use, helping organizations scale trusted AI, reduce duplicated logic and unnecessary costs, and keep business definitions consistent as technology changes. 35 years Run at Fortune 500 scale 35 years in production at enterprise scale. Pfizer runs 15,000 users across 27 markets on Mosaic. So do companies like Fannie Mae and Porsche. 53% fewer tokens Cost savings you can measure Mosaic serves repeat queries from cache and sends AI compressed context instead of raw data, so cloud and token costs drop. Up to 53% fewer tokens used with the Mosaic semantic layer. 100% accuracy Give AI your definitions and it stops making them up Give AI your governed context and it stops guessing. In our benchmark, Mosaic answered with 100% accuracy, against 88.2% querying raw data. Why AI fails Hallucinations AI invents definitions it doesn't know. If the model doesn't have access to governed context or verified data, it might give answers that sound right but aren't. Inconsistent Answers The same request returns different information every time it's asked. When the model writes its own query, you get different numbers from the same data. Ungoverned AI Without access controls applied to it, an agent can surface a record to the wrong person or expose sensitive fields. What changes when you add Strategy Mosaic Mosaic is a universal semantic layer that sits between your data and your tools. It turns raw tables into governed business definitions that every model and tool shares. Mosaic gives every model one governed definition of revenue, a customer, or churn to work from, so it answers from verified logic instead of inventing it. The model handles the language. Mosaic handles the query. Its deterministic SQL engine means the same question returns the same result for every model, every team, and every tool. Every query runs through your access controls before it returns anything. Sentinel logs who asked, what data it reached, and what it cost. Govern your definitions once. Use them everywhere. Mosaic is a universal semantic layer that sits between your data sources and your tools. It turns raw database tables into certified, governed business concepts that every tool and agent shares. The model handles the language. Mosaic handles the query. Unlike a model's probabilistic guesswork, Mosaic's deterministic SQL engine means the same question always produces the same query, and the same answer, every time, for every model and every user. It works with your stack, not against it Mosaic sits above your data sources and beneath your tools, so nothing moves or migrates. Bring any AI: Claude, Copilot, Gemini, or any MCP agent. Run any cloud or warehouse, or several at once, and your definitions don't change. 200+ native connectors reach every major cloud, warehouse, and lakehouse. Mosaic automatically caches repeated queries across AI agents, BI tools, and LLM applications, reducing warehouse compute costs while keeping definitions consistent everywhere. Run by companies that cannot afford a wrong answer Start Your Mosaic Journey On consumption-based cloud contracts, Mosaic deflects redundant queries away from the warehouse. See your stack-specific savings estimate. One governed entry point for every AI tool through Mosaic's MCP server. Approve the architecture once. Every future agent uses the same path. Connect your warehouse, model your first use case, and see governed answers in your own tools within weeks. Frequently Asked Questions