AI agents transform the industry by automating workflows, increasing productivity and allowing intelligent decision making. Businesses use AI agents to process insurance claims, IT service tables, optimize the logistics of the supplier chain and even help healthcare professionals to analyze medical records. The potential is huge and we are excited that we can introduce two powerful innovations in Azure AI Foundry.
AI agents transform the industry by automating workflows, increasing productivity and allowing intelligent decision making. Businesses use AI agents to process insurance claims, IT service tables, optimize the logistics of the supplier chain and even help healthcare professionals to analyze medical records. The potential is huge and we are excited to introduce two powerful innovations in Azure AI Foundry:
- API Answer: Powerful APIs allowing A-Powere applications to retain information, data data and act without problems.
- Agent using a computer (Cua): AI breakthrough model that navigates the software interface, performs tasks and automates workflows.
Together, these skills seize business to reimagine AI not only as an assistant – your active digital workforce. Customers of the company will soon gain access to these innovations that control automation, efficiency and intelligence on a scale.
Improvement of AI agents using API responses
The API responses is the key to unlocking AI agent in Azure AI founders and transforms how businesses use AI for impact in the real world. It is a new basis for the use of powerful Azure Openai Service tools that combines the simplicity of chat interface with the completion of the chat with advanced abilities that can be used using the API and Azure AI agents. API responses allows trouble -free interaction with tools such as cua, function calls and file search – all in one API call. This API interface allows AI systems to load data, process information and take steps – SSEAMESSEDSE TONCESING Agent AI with business workflows.
How does the API responses work
The API reaction provides a structured responsible format that allows AI to interact with multiple tools and attach a context across interactions. Supports:
- Call tool in one simple call API: Now developers can easily integrate AI tools, which will be more efficient.
- Calculate: Use a computer to use a computer in API responses to control automation and performing software interactions.
- File Search: Dynamically interact with business data data and extract information about reporting.
- Calling a function: Develop and induce your own features to increase AI capacity.
- Chaining of responses to conversations: Observe the interactions by joining responses together using ID of a single response and ensuring continuity in controlled dialogues A.
- Data Privacy Personal Data Protection Enterprise-Grade: Built with trusted standards of security and compliance with Azure regulations and ensuring data protection for organizations.
By consolidating, thinking and performing actions into a single API API, API simplifies the development of AI, thus reducing the complexity of orchestracting multiple AI tools within the automation pipe.
This scalability is suitable for enterprise use of ACE across industries such as customer service, IT operations, finance and supply chain management, where the authorization AI can streamline workflows and improve efficiency. For even greater flexibility and control, organizations can explore the Azure AI agent, which offers other tools and models for the development and scaling of AI agents. The Azure AI Agent is integrated with a semantic core and autogen, allowing a smooth multi-agent orchestration for more complex scenarios that require more agents to work together on tasks.
Strengthening AI agents with an agent using a computer
A computer agent (CUA) is a specialized AI model on Azure Openi, which allows AI to interact with graphical user interfaces (GUI), navigate applications and multi-stage tasks-via natural language instructions. Unlike traditional automation tools that rely on predefined scripts or API -based integration, CUA can interpret visual elements, dynamically adapt, and take steps based on the content on the screen.
What is doing a computer using a unique agent?
- Automatic user interface navigation: Can open the app, click on buttons, fill in forms and multi -page workflows.
- Dynamic adaptation: It interprets the user interface changes and adjusts the actions accordingly, reducing the relying on solid automation scripts.
- Carrying out tasks of cross application: Operates across web and desktop applications and releases the aforementioned systems disappear without API addictions.
- The interface of the command interface of the natural language: Users can describe the task in a simple language and CUA determines the correct interactions of the user interface to perform.
Thanks to today’s announcement, developers can start building further agent capabilities. Since businesses are trying to deploy this technology on a scale, we evaluate integration with the Windows 365 and Azure virtual computer to allow CUA automation to run smoothly on cloud computers or virtual machines (VMS).
Ensuring secure and credible automation AI
Since AI systems become more autonomous, critical security, disturbability and alignment with the human intention. The CUA is one of the first AI agent models that are able to directly interact with software and bring new challenges in preventing abuse, unintended actions and contradictory risks. To deal with this, Microsoft and Openai have implemented a multilayer safety approach that includes model, system and deployment level.
The CUA is developed with warranty to refuse harmful tasks, rejected unauthorized measures and prevents abuse. At the system level, Microsoft implements the filtering and monitoring of the corporate level to help detect and prevent policy violations. To minimize unintended actions, CUA is designed to ask the user to confirm the user before performing irreversible tasks and reduce high -risk actions such as financial transactions.
Microsoft’s trusted AI framework also ensures observability, logging and audit compliance with the company’s deployment regulations. Automated and human detection systems monitor the formulas of implementation, identify unusual behavior and promotion of management policy. These guarantees are continuously refined based on internal red creation, external audits and real -world testing to strengthen protection against rapid injections, curtain manipulations and unauthorized approach. Due to the current level of the reliability of the Cua-but, especially in non-human supervision of human supervision, it remains strongly recommended for sensitive operations.
As AI agents develop, Microsoft has committed itself to transparency, security and continued risk alleviation. By combining Birtet-in Cua with Azure’s Enterprise Comprise and Governce Tools, the organization can deploy automation of Ai-sink with certainty, to accept safe and responsible AI on a scale.
We start with Cua and API API
The Azure AI Foundry continues to shift the AI-Powred automation. Enterprise customers will receive access to API and CUA API and CUA responses in the coming weeks.
We are enthusiastic about how developers and business innovate with these new abilities.