Microsoft infuses corporate agents with deep reasoning, reveals an agent of data analyst that outperforms competitors
Join our daily and weekly newsletters for the latest updates and exclusive content of a leading AI coverage industry. Learn more
Microsoft have Build the biggest AI Agent ecosystemAnd now it is expanding its leading role with powerful new opportunities that position the company forward in one of Enterprise Tech’s most exciting segments.
The company announced on Tuesday evening two significant additions to its Copilot Studio platform: Deep reasoning opportunities that allow agents to deal with complex problems through careful, methodical thinking and flows of agents that combine AI’s flexibility with determinated automation of business processes. Microsoft also introduced two specialized agents for deep reasoning for Microsoft 365 Copilot: researcher and analyzer.
“We already have clients with thousands of agents,” Microsoft’s Corporate Vice President for Business and Industry Copilot Charles Lamanna, before Venturebeat in an exclusive interview on Monday. “You are starting to have this type of agent work force, in which no matter what the job is, you probably have an agent that can help you do it faster.”
The Microsoft Analyzer’s distinctive agent
While the researcher’s agent reflects the capabilities of competitors such as the Openai and Deep Research Deep Research, the Microsoft analyst agent represents a more niforted supply. Designed to function as a personal data scientist, the analyzer agent can process various data sources, including Excel, CSV files and built -in tables in documents that generate insights by executing and visualization of code.
“This is not a major model outside the shelf,” Lamana stressed. “This is quite a few extensions and setup and training at the top of the basic models.” Microsoft uses his deep understanding of Excel’s work flows and data analysis models to create an agent to be aligned with the way the corporation users actually work with data.
The analyst can automatically generate PYTHON code for processing uploaded data files, producing visualizations and providing business insights without requiring technical expertise from users. This makes it particularly valuable for financial analysis, budget forecasting and cases of use of operational reporting, which usually require extensive data preparation.
Deep Reflections: Carrying Critical Thinking to Enterprise Agents
The ability to think of Microsoft’s deep reasoning extends the capabilities of agents beyond the simple execution of tasks to complex judgment and analytical work. By integrating models for sophisticated reflections such as Openai’s O1 and connecting them to business data, these agents can handle ambiguous business problems more methodically.
The system dynamically determines when to refer to deeper reasoning, or implicitly based on the complexity of the task or explicitly, when users include prompts such as “reason over this” or “think really hard for it”. Behind the scenes, the platform analyzes the instructions, evaluates the context and selects appropriate tools based on the requirements of the task.
This allows scenarios that were previously difficult to automate. For example, a large telecommunication company uses agents for deep reasoning to generate complex RFPs responses, assembling information from multiple internal documents and sources of knowledge, Laman told VentureBeat. Similarly, Thomson Reuters uses these capabilities for proper verification in mergers and examinations for acquisition, processing unstructured documents to identify insights, he said. See an example of the agent’s reasoning for the video in the video below:
Agent Flows: Automation of Rethinking Processes
Microsoft has also introduced streams of agents that effectively develop robotic processes (RPA), combining work processes based on rules with AI reasoning. This addresses the client’s requirements for integrating deterministic business logic with flexible AI capabilities.
“Sometimes they don’t want the model to be free. They don’t want to make their own decisions. They want to have firmly encoded business rules,” Lamana explained. “Another time they want the agent to deal with freestyle and make court calls.”
This hybrid approach enables scenarios such as intelligent fraud prevention in which the agent’s flow can use conditional logic to route requests to restore a higher value to AI agent for deep analysis against political documents.
Pets at home, based in the UK pet supplier, has already deployed this technology to prevent fraud. Lamana revealed that the company saved “over a million pounds” through implementation. Similarly, Dow Chemical realized “millions of dollars saved for transportation and load management” through optimization based on agents.
Below is a video showing the agent’s flows at work:
Advantage of Microsoft Graph
Central to the Microsoft agents strategy is its integration of enterprise data through Microsoft Graph, which is a complete mapping of workplace relationships between people, documents, emails, calendar events and business data. This provides the agents with contextual awareness that the generic models are missing.
“The less known secret capacity of Microsoft’s graphics is that we are able to improve the relevance of the graphics based on commitment and how closely it is related to some files,” Laman revealed. The system identifies which documents are the most referenced, shared or commented, ensuring that agents refer to authoritative sources, not outdated copies.
This approach gives Microsoft a significantly competitive advantage over the autonomous AI suppliers. While competitors can offer modern models, Microsoft combines those with workplace context and fine setting, optimized explicitly for cases of use of Microsoft businesses and tools.
Microsoft can use the same web data and model technology that competitors can, Lamana noted, “But then we have all the content in the enterprise.” This creates the effect of a flywheel in which any new interaction of the agent further enriches the understanding of the graphics for the models in the workplace.
Acceptance and accessibility of enterprise
Microsoft is a priority to make these powerful opportunities accessible to organizations with various technical resources, Lamana said. The agents are exposed directly to Copilot, which allows users to interact through natural language without quick engineering experience.
Meanwhile, Copilot Studio provides a low -custom development environment. “It is in our DNA to have a tool for everyone, not just for people who can start Python SDK and call, but anyone can start building these agents,” Laman emphasized.
This accessibility approach nourishes quickly. Early Microsoft revealed that over 100,000 organizations used Copilot Studio and that Over 400,000 agents have been created in the last quarterS
The competitive landscape
While Microsoft seems to be developing today the introduction of the Enterprise agent, competition is increasing. Google has expanded its twin capabilities Agents and Agriculturewhile Openai SDK’s model and agents provide powerful reasoning and agent instruments for developerS Large company applications such as Salesforce, Oracle, Servicenow, SAP and others have launched agency platforms for their customers in the last year. And also on Tuesday AWS on Amazon released an AI agent called Amazon Q in Quicksight to let employees engage in natural language to perform data analysis without specialized skills.
Employees can use natural language to perform expert level data analysis, ask questions about what-if they get recommendations that can be reflected, help them unlock new insights and make decisions faster
However, the advantage of Microsoft is in its more comprehensive approach-a strong connection with the leading company for reasoning, Openai, while offering a model choice, corporate class infrastructure, extensive integration of workplace tools and focus on business results, not Raw AIs. Microsoft has created an ecosystem that looks like the best practice by combining personal copywriting that understands individual models of working with specialized agents for specific business processes.
For persons making decisions of the enterprise, the message is clear: agency technology has aged beyond experimentation with practical business applications with a measurable return on investment. The choice of platform is increasingly dependent on integration with existing tools and data. In this area, Microsoft has an advantage in many areas of application due to the number of users, for example, in Excel and Power Automate.
Watch my full interview with Charles Lamana, embedded in the first hand, to hear how Microsoft manages your agent strategy, what these new opportunities for business users mean, and how organizations use agents to achieve measurable business results:
English