Blog

Maximilian Kenfenheuer

Manager · Microsoft 365 & AI Transformation @ BearingPoint

Featured

#Microsoft365Copilot #MicrosoftAI #Copilot #AITransformation #FutureOfWork

Measuring Completed Work: Baselines, Assisted Hours and Business Outcomes

Thursday, September 10, 2026 · Maximilian Kenfenheuer

A team can generate thousands of prompts without improving a single core process. Another uses AI rarely but saves hours in contract review. Microsoft increasingly measures Copilot value by completed work and assisted hours instead of activity. Its internal guidance recommends baselines for painful processes before the rollout. In this article, I describe a three-layer measurement model of adoption signals, time and effort indicators and business outcomes, and explain why role-based deployment makes value measurable.

Power BI Grounding: Copilot Reasons Over Semantic Models

Wednesday, September 9, 2026 · Maximilian Kenfenheuer

Since the August 2026 update, Microsoft 365 Copilot reasons over Power BI reports and semantic models in Chat and Cowork. Copilot thus works with governed business definitions: how revenue is defined, how margin is calculated, which KPIs leadership uses and who may see which numbers. At the same time, a poorly defined KPI now spreads faster. In this article, I explain why semantic models become the foundation for trustworthy AI answers and how organizations can prepare their analytics layer.

From Draft to Deliverable: Agentic Copilot in the Core Apps

Saturday, September 5, 2026 · Maximilian Kenfenheuer

On April 22, Microsoft made agentic capabilities in Word, Excel and PowerPoint generally available and brought business apps into Copilot via agents. AI moves to where proposals are written, financial models built and board decks assembled. At the same time, human judgment gains weight: a polished deck is not automatically a sound strategy, and a well-structured spreadsheet is not automatically a correct model. In this article, I explain what changes when AI participates in the work and which five readiness questions organizations should answer.

Posts tagged #CopilotStudio

Frontier Tuning: Reinforcement Learning Inside the Compliance Boundary

Friday, August 21, 2026 · Maximilian Kenfenheuer

#Microsoft365Copilot #MicrosoftAI #CopilotStudio #EnterpriseAI #AITransformation

Microsoft introduced Frontier Tuning at Build. It applies reinforcement learning inside the organization's compliance boundary, using its own data, conventions and workflows. The results are tuned models, skills, orchestration logic and a runtime harness that inherit existing access policies. Frontier Tuning comes to Copilot Studio and Microsoft Foundry. Pearson, EY, Bristol Myers Squibb and McKinsey are among the early users. In this article, I explain how tuning differs from prompting and retrieval and which questions organizations should answer before tuning agents.

Work IQ as Context Infrastructure for Copilot and Agents

Monday, August 17, 2026 · Maximilian Kenfenheuer

#MicrosoftAI #Microsoft365Copilot #CopilotStudio #EnterpriseAI #AITransformation

For many organizations, access to a capable model is no longer the problem. The problem is whether the model has enough business context to be reliably useful. Work IQ turns signals from email, meetings, files, chats and business systems into context for Copilot and agents. It works permission-aware and user-scoped, with central logging and policy enforcement. In this article, I explain why this infrastructure layer may matter more than the next interface feature and what organizations should prepare in information architecture and governance.

Copilot Tuning and Multi-Agent Orchestration: Adapting AI to the Organization

Monday, August 10, 2026 · Maximilian Kenfenheuer

#Microsoft365Copilot #MicrosoftAI #CopilotStudio #EnterpriseAI #AITransformation

A law firm has its own drafting style, a consulting company its own industry playbooks. General-purpose models do not know either. With Microsoft 365 Copilot Tuning, organizations tune models with their own data, workflows and processes in Copilot Studio, low-code and without a data science team. Multi-agent orchestration lets HR, IT and marketing agents share work, for example during onboarding. In this article, I explain why organizational fit may become a stronger advantage than access to the most powerful general model.

Model Choice in Copilot Cowork: From AI Access to a Model Portfolio

Saturday, August 8, 2026 · Maximilian Kenfenheuer

#Microsoft365Copilot #MicrosoftAI #AI #CopilotStudio #EnterpriseAI

Copilot Cowork now offers several models, depending on what the organization allows: Auto, Claude Sonnet 5, Claude Opus 4.8, GPT 5.5 (Frontier) and a paired Sonnet + Opus Advisor mode. A quick meeting summary needs different capabilities than a high-stakes research synthesis. Microsoft makes this choice available inside enterprise controls, with visible retention notices and admin settings per model family. In this article, I explain why enterprises move from AI access to managing a model portfolio and what IT leaders should watch.

Data, Memory, Inference: The Three Layers of Work IQ

Thursday, August 6, 2026 · Maximilian Kenfenheuer

#Microsoft365Copilot #MicrosoftAI #Copilot #AI #EnterpriseAI #CopilotStudio

An excellent model with thin enterprise context still produces generic output. The advantage lies with the platform that provides context. Work IQ builds this context in three layers: data from Microsoft 365 and business systems, memory about how teams work, and inference for reasoning and action. Microsoft offers it via REST, agent-to-agent and a remote MCP server, on a consumption basis and independent of Copilot licensing. In this article, I explain why this context infrastructure may become more durable than any assistant interface.