Amazon Nova Act - Amazon's general-purpose AI agent that autonomously performs web page tasks.

What is the Amazon Nova Act?

Amazon Nova Act is a general-purpose AI agent from Amazon AGI Labs, used to perform tasks within web browsers. Amazon Nova Act allows developers to build agent application prototypes based on the accompanying SDK, performing actions such as submitting leave requests, scheduling calendar events, or sending automated email replies. Amazon Nova Act improves the reliability and efficiency of task execution by breaking down complex tasks into atomic commands and directly manipulating the browser using Playwright. Amazon Nova Act helps developers build agent applications that automate network operations, and in the future, it is expected to enable more complex, multi-step tasks.

Main functions of Amazon Nova Act

• Decompose complex tasksIt supports developers in breaking down complex, multi-step tasks into reliable atomic commands, such as searching, filling out forms, and clicking buttons.
• Add detailed instructionsIt supports developers in adding more detailed instructions to each command, such as "Do not accept insurance sales" or "Select a specific date," to ensure that tasks are executed as expected.
• Supports multi-tasking operationIt can handle multiple tasks simultaneously, and implement parallel processing of tasks based on Python code to improve efficiency.
• Provide SDK toolkitAmazon has launched a software development kit (SDK) to help developers build prototypes of intelligent agent applications.

The technical principles of Amazon Nova Act

• Natural Language Processing (NLP)Based on a pre-trained language model, it parses user intent and translates it into specific task operations. NLP technology supports multilingual processing, enabling Amazon Nova Act to serve users of different languages worldwide.
• Web page automation technologyThis feature simulates human web browsing behavior using automation tools (such as Selenium or Puppeteer). It automatically fills in user-provided data by parsing the HTML structure of web pages and identifying form fields. It also supports extracting necessary information from web pages, such as restaurant reservation confirmation numbers and order status.
• Machine Learning and Reinforcement LearningMachine learning algorithms optimize the efficiency and accuracy of task execution. Based on a large amount of training data, the model can learn the optimal task execution path. Learning techniques enable Nova Act to self-optimize based on feedback while performing tasks. For example, if an error is encountered when booking a restaurant, reinforcement learning is used to adjust the strategy and avoid repeating the same mistake.
• API calls and integrationIt interacts with various external services (such as restaurant reservation systems and online shopping platforms) via API calls to complete complex tasks. Amazon Nova Act provides an SDK, allowing developers to create more customized applications based on their own business logic.

Amazon Nova Act project address

• Project official website:https://labs.amazon.science/blog/nova-act

Application scenarios of Amazon Nova Act

• Office AutomationAutomatically submit leave requests, schedule meetings, fill out forms, and generate reports.
• e-commerceIt can automatically complete online shopping, regular procurement, and price monitoring.
• Personal life managementAutomatically manages schedules, healthy living plans, and restaurant reservations.
• Complex task processingAutomated organization of events (such as weddings and parties) and execution of complex IT tasks.
• Enterprise business processesAutomated handling of customer relationships, supply chain management, and data analysis.