DNA: Enterprise AI Comes to Life in Just Two Weeks

Leading telecom operator DNA and Codento proved that Google Agentspace delivers in the world of agentic AI.

DNA Aims to Take the Lead in Agentic AI

Telecommunications operator DNA, part of Telenor Group, has set itself a target to become an AI-first, cloud native company. DNA has adopted a forward-thinking mindset by piloting new AI technologies in its own operations. Once validated, the company extends these innovations to its clients, enabling them to leverage cutting-edge solutions with confidence. As part of the strategic initiative to remain at the vanguard of AI innovation, DNA decided to pilot Google Agentspace and become part of an elite group of global customers who have had the chance to do so.

Codento, as DNA’s and Google Cloud’s trusted partner, supported DNA’s experienced AI teams in doing just that. The aim was to discover how the brand new Agentspace platform can integrate into DNA’s B2B business data to drive measurable results and to unlock tangible business value.

Google Agentspace is a novel enterprise AI platform powered by some of Google Cloud’s most advanced technologies. First, it includes easy-to-use no-code tools for creating expert AI agents that automate complex tasks and seamlessly integrate with enterprise tools securely and at scale. Second, Agentspace brings Google-quality search to the workplace, helping employees instantly find information across all tools and systems. Finally, Google’s NotebookLM turns any content into dynamic, AI-powered interactions, explaining, visualizing, and contextualizing data across formats.

Google Agentspace in Action

While Agentspace is not generally available yet, Codento, a leading Google Cloud partner, has introduced the “AI Agent Launchpad” program that enables its customers to test the full Agentspace functionality in a sandbox environment.

Jointly with DNA, Codento agreed on a two-week Proof of Concept based on the Launchpad concept to test three use cases:

A personalized sales lead assistant to equip sales representatives with company-specific insights, tailored arguments, and relevant references—streamlining call preparation, highlighting key pain points, and supporting more consistent, effective sales engagement.
An agreement analysis agent to automatically compare B2B service agreements against templates, highlight deviations, and answer key questions, reducing manual review effort, supporting standardization, and potentially saving days of work.
An intranet chatbot to search internal knowledge from SharePoint-based service descriptions, aiming to reduce customer service handling time, improve productivity, and benchmark the performance against DNA’s existing AI solution based on Retrieval-Augmented Generation (RAG).

Pioneering Google Cloud Partner Codento Enables Success

Specialized in Google Cloud, Codento has been DNA’s trusted AI journey partner. They have helped identify the right course of action; first in supporting DNA build its fresh Security Operations Center offering on top of Google Cloud AI and, now, experimenting on selected Agentspace use cases.

Codento’s AI Agent Launchpad program offered DNA a unique, targeted, hands-on experience with Google Agentspace. With its case-led approach, the program helped develop and distill DNA’s ideas and turn them into fully fledged Agentspace applications that DNA can test, evaluate, and use without committing to purchasing licenses.

Delivering Results on Agentspace in Just Two Weeks

DNA’s seasoned AI experts who participated in the Launchpad program found Google Agentspace fast and intuitive to set up, with one user calling five-minute file indexing via NotebookLM “almost quite incredible.” Testers praised the platform’s user-friendly interface, structured search results, and the ease of building and iterating on AI agents, making it ideal for rapid proof-of-concept development.

Across use cases, the testers considered Agentspace to be accurate, easy to learn, and effective at delivering concise, data-grounded answers. While full integration with Salesforce wasn’t possible during the two-week sandbox trial, the personalized sales lead assistant still impressed with its clarity and speed. The agreement assistant successfully identified deviations from templates, and the intranet chatbot provided “concise and straight-to-the-point answers.” A key takeaway for testers was the importance of precise prompting, as even small language changes could affect results.

Overall, the PoC demonstrated Agentspace’s strong potential, giving DNA and Codento confidence to continue exploring its use across business units.

In DNA’s Own Words

DNA’s Vice President of Customer Operations, Lasse Huttunen, summarizes the findings of these pioneering experiments on Google Agentspace into three points:

“First, the Agentspace platform was surprisingly easy to implement and versatile. Second, we must remember that AI agents are pretty sensitive. That means we must be precise when communicating with them and ensure our data is available. Third, based on this short experience, our AI specialists are super excited about the opportunities. Agentspace proved to be a very viable platform for AI change at DNA.”

The technologies we used:

Google technology	Usage in the DNA case
Google Agentspace	Gemini 2.0 queries and synthesizes unstructured data from cloud storage and grounds it via Google’s search infrastructure with Internet-hosted assets. Insights are delivered through a conversational, assistive AI interface.