- The new tools of researchers and analysts are here for Microsoft 365 Copilot
- The new deep reasoning tool in Copilot Studio will use models like Openi's O1
- It will take time to analyze and recognize patterns to make better decisions.
Microsoft has announced a series of updates to its generative co -driver tool to take advantage of the latest innovations in AI, including Openai's most powerful models, as well as two new agents designed for work.
The company says that its new tools of researchers and analysts will bring advanced reasoning with co -pilots to daily work, which makes employees more productive and effective.
“They analyze large amounts of information with safe access and according to their work data (their emails, meetings, files, chats and more, and the web to offer highly qualified experience on request,” a Microsoft 365 blog post announces the new versions.
Researcher and analyst at Microsoft 365 Copilot
Microsoft says that the new researchers will seek to deliver ideas, “with greater quality and precision of the possible.”
It will use OpenAI's deep research model with the advanced co -ilot orchestration of Microsoft 365 and deep search capabilities to address complex or multiple steps tasks, such as building a market sales strategy, or building a quarterly report based on all your work data.
The researcher can also connect to third -party data to provide even more complete information, with Salesforce, Servicenow, Confluence and more feeding directly in Microsoft 365 Copilot, as well as other company agents, such as the recently announced sales chat.
Built in the OPENAI O3-MINI Reasoning Model, the analyst focuses on advanced data analysis tasks, using thought chain and python reasoning to help convert the unprocessed data into demand forecasts, visualizations of trends and income projections.
Deep reasoning reaches Microsoft Copilot
Taking LinkedIn to announce the changes, the CVP of Copilot Business and Industrial Charles Lamanna introduced a deep reasoning in Copilot Studio, which uses advanced reasoning models such as Openi's O1 for tasks that require a detailed analysis, methodical thinking and nuanced understanding.
Lamanna highlighted how these models take longer to analyze complex data sets, recognize complex patterns and make reflexive decisions, which makes them ideal for addressing complex problems.
Copilot takes into account the analysis of inputs and instructions to decide when it is appropriate to spend the extra time taking advantage of deep reasoning, but Lamonna also explained that agents creators can choose to include the key word 'reason' to invoke deep reasoning as part of a user's message.
At the same time, Microsoft lifted the enveloping agent flows, which are designed to connect the holes where AI agents have shown that they are not so capable. “Between agents and agent flows, it is possible to automate any task I can imagine,” added Lamonna.
Agents flows are designed for repetitive tasks such as document processing, financial approvals and compliance tasks, and can operate independently or as skills within existing agents.
In an example, Lamonna describes how an agent flow can direct customer comments to a comment monitoring agent for summary and the creation of action points, or a customer service agent to answer customer questions and take measures to solve the problem.
The final piece of the puzzle will allow companies to react to signals automatically through 50 pre -constructed triggers: they are called autonomous agents and promise to improve the efficiency in the acquisition of devices, the discovery of suppliers, the prevention of fraud and other knowledge works.
The deep reasoning in Copilot Studio is available in today's preview, with the flows of agents that will generally be available on March 31.
