Google continues to work on its big bet on AI, both for people and businesses. In a recent development, Google has launched Agentspace, a suite of AI-powered agents for enterprises powered by Google’s Gemini LLMs.
Agentspace brings the power of Google’s AI and search tech to enterprises
Many people have successfully incorporated AI-powered chatbots into their daily lives by now. However, there are also more specialized, autonomous, and proactive developments, known as AI agents. These “assistants” are initially targeting companies that aim to enhance employee productivity. They are ideal for optimizing and focusing on specific tasks. Google Agentspace is the new AI agent platform from the Mountain View giant for enterprises.
According to Google, company employees use between four and six tools to solve a single question or task on average. One of the main benefits of artificial intelligence is process optimization, so it seems ideal to solve this problem. Google designed Agentspace with that in mind.
Google’s AI agents combine Gemini’s “reasoning” capabilities, the power of Google’s search engine, and the data of enterprise clients. The result is highly effective assistants that can not only solve queries but also execute complex tasks with a few prompts.
NotebookLM Plus and multimodal search are key parts of the suite
Google Agentspace is made up of three main features. NotebookLM Plus, an enterprise version of the Mountain View giant’s research tool, is one of them. NotebookLM Plus offers the same options as the NotebookLM we already know. That is, it can analyze content to generate summaries, potential new ideas, or even start a “debate” on a topic between AI-powered hosts, among other things.
However, the NotebookLM Plus boasts more security and privacy features, something essential for companies. Google is also implementing the new Gemini 2.0 Flash model in NotebookLM, so it will be more powerful and efficient than ever.
Multimodal search is the second key element of Google Agentspace. This gives agents advanced search capabilities—powered by Google AI models—based on enterprise data and knowledge. The best thing about the feature is that it works with both structured data (documents, emails, etc.) and unstructured data (such as tables). That means that every bit of info will be useful to boost company productivity.
On top of that, Google has integrated its translation engine, so multimodal search will be useful for company data in almost any language. There are connectors for multiple enterprise content management platforms, including Confluence, Google Drive, Jira, Microsoft SharePoint, ServiceNow, among others.
All custom AI agents accessible to employees from one place
The third key element of Google Agentspace is a hub from where employees can access all AI-powered agents trained on the company’s data. This will allow them to easily find the most suitable agent for the task they need to complete. Google is even working on a visual coding tool so that employees can create their own AI agents specialized in a task. This will be very useful for those who find that the AI agents provided by the company are not enough, or for those who feel that they can further optimize their work on their own.
According to Google, its AI agents can facilitate and automate multiple tasks. The company provides examples, such as the ability for business analysts to easily identify ongoing trends in the industry. HR teams can turn to AI agents for better and faster onboarding. Meanwhile, software engineers can find and resolve bugs in the code more easily. The possibilities are endless.
Lastly, Google Agentspace is designed on a secure infrastructure for enterprises. It offers features such as “role-based access control (RBAC), VPC service controls, and IAM integration.” This ensures that the company can isolate sensitive data only to authorized employees or prevent potential information leaks, for example.
