Automated up to 80% of transactional decision-making and communication
Improved average customer response time potentially from 42 hours to near real time for a significant volume of orders
Reduced processing time by 50%
Consolidated user workflow from up to five different systems into a single interface
Prevented churn and accelerated onboarding for new hires and increased engagement among younger employees
Danfoss partnered with Go Autonomous to build a solution on Google Cloud to automate email-based B2B order processing across its global operations, helping reduce turnaround time, manual workloads, and improve user experience.
Danfoss partnered with Go Autonomous to build a solution on Google Cloud to automate email-based B2B order processing across its global operations, helping reduce turnaround time, manual workloads, and improve user experience.
Closing customer service gaps with Autonomous Commerce
While conversations about the fourth Industrial Revolution tend to focus on the shop floor, Danfoss is driving innovation in a less obvious, but equally critical space: the inbox. A global manufacturer of energy-efficient solutions, Danfoss operates in over 100 countries, with a broad portfolio comprising original equipment manufacturers, distributors, and end users. That scale brings complexity and with it, a consistently high volume of customer emails.
"We have well-established digital systems for business transactions with larger partners," explains Carlos García, Head of Digital Business at Danfoss. "But there was a gap when it came to businesses and customers with more complex ordering needs." The company's digital ordering tool for customers outside its existing system-to-system integrations was falling short of expectations. "It often required as much manual input as email-based ordering," García says.
In order to close this gap, reduce the customer service team's manual workload, and enable faster, more efficient order handling at scale, Danfoss discovered the software provider Go Autonomous through Google Marketplace and partnered with them to implement their Autonomous Commerce solution. The platform automates B2B transactions, helping to streamline workflows and reduce manual input.
"The fact that Go Autonomous was hosted on Google Cloud, an infrastructure we already trust and use ourselves, made the choice even easier," states García. "We needed an AI-first, secure, and compliant platform, and this partnership delivered."
Potentially reducing order processing time from 42 hours to under a minute, powered by the Google Kubernetes Engine
Previously, the need for semi-automatic processing meant that order handling could take hours — or even days — due to the need for human intervention across multiple systems. For large-scale manufacturing companies like Danfoss, slow workflows tie up capital. The new platform, however, enables real-time handling of orders, shortening the cash conversion cycle and is expected to save millions of euros for Danfoss. "Our turnaround time for manual orders used to be around 42 hours. Now we've cut potentially the order entry processing time down dramatically, and we're looking to do real-time responses for a significant part of our incoming orders," García says. "In the near future, a buyer can send an order and receive an acknowledgement within a minute."
The Autonomous Commerce platform is powered by specialized AI models, trained on millions of B2B transactions from across the world. The solution uses advanced machine learning models and AI agents to interpret and execute complex customer requests from incoming emails, such as orders or quotes, by triggering the appropriate actions in Danfoss's enterprise resource planning (ERP) software.
Choosing a vertical platform for handling transactional communication was a critical factor for Danfoss. "The more horizontal AI platforms couldn't scale to our needs," explains García. "Our goal was twofold: to make as many transactions fully autonomous, while ensuring that the process is seamless when a human decision is needed. This requires a user interface that deeply understands the context of our industry and business workflow. We refused to settle for a solution that would permanently rely on a human-in-the-loop, as our ambition was always to go fully autonomous with as many orders as possible."
This approach of integrating with a purpose-built platform was necessary because building it based on a more horizontal solution with a standard large language model (LLM) was unsuitable, given the sheer variability of B2B communications. Customer emails often contain a mix of free text, tables, and attachments in multiple formats, like PDF, images, and spreadsheets. Handling such B2B transactions requires domain-specific artificial intelligence to understand details like pricing mismatches, missing part numbers, and Danfoss-specific business workflows — challenges for which the more horizontal AI platforms are not built.
To address this, Danfoss collaborated closely with the customer success team at Go Autonomous. Together, they configured the Autonomous Commerce platform and developed ERP and CRM connectors, tailoring the solution to Danfoss's precise business workflows. To deliver the enterprise-grade availability required by Danfoss, Go Autonomous hosts its proprietary AI architecture on Google Cloud, using Google Kubernetes Engine (GKE) to manage and scale its platform reliably.
"Google Cloud provides both excellent training compute and inference capabilities. As far as we're concerned, this makes Google the ideal hosting and inference partner," says Bjarke Ruse Sejersen, Founder and CEO of Go Autonomous.
Google Cloud provides both excellent training compute and inference capabilities. As far as we're concerned, this makes Google the ideal hosting and inference partner.
Bjarke Ruse Sejersen
Founder and CEO, Go Autonomous
Helping sales staff focus on what they do best
Since automating order management with the Autonomous Commerce platform, the Danfoss team estimates that average time savings per order are around five minutes, as well as the number of orders needing to be processed fully manual is decreasing steadily. "Nowadays, above 80% of those decisions are made by an AI agent, enabling staff to focus on more impactful value-added activities," García says. "We tend to look at AI as a cost-saving tool, but there's also a growth dimension, the value of freeing up mental bandwidth. People who constantly make decisions — big or small — can get overloaded, and that can affect decision quality."
The Go Autonomous UI is purpose-built for handling B2B transactions and communication, and was configured in consultation with the customers and employees who would be using it every day. Sales and customer support personnel now work with a single user interface, compared to five different systems previously. The platform's modular architecture is deployed as containerized services on Google Cloud infrastructure, and fully integrated with the company's enterprise resource planning system and customer relationship management software. As a result, new hires often don't even realize they are working across different solutions.
"These systems are incredibly complex, and it typically takes months to onboard someone," García says. "We've simplified the user experience so much that onboarding is much faster, the job is much more enjoyable, and churn is reduced." The latest rollouts have been met with enthusiasm from staff. "It's not something we initially tracked, but having a solution like this really helps with engagement and retention," García adds. "We're getting great feedback from users daily, which is really nice to see — and we're not stopping here."
Turning transactional data into actionable insights
At this stage, Danfoss staff are still required to address more complex customer transactions, such as pricing exceptions. Over the next few months, the company is looking forward to working on connecting the Go Autonomous platform with BigQuery, a unified data platform, to drive data-informed decision-making that could enable our AI agents to handle those as well.
"We now have a platform that helps us make sense of all the unstructured data we receive in our inboxes, such as free text, images, spreadsheets, and more," García says. "It would be a natural next step to analyze that data to derive actionable insights. We're looking at connecting to BigQuery because it would require analysis of massive datasets. One of the reasons Google Cloud has been such a strong partner is that the infrastructure is among the best for supporting these kinds of high-performance use cases."
We're looking at connecting BigQuery because it would require high-speed analysis of massive datasets. One of the reasons Google Cloud has been such a strong partner is that the infrastructure is among the best for supporting these kinds of high-performance use cases.
Carlos García
Head of Digital Business, Danfoss
The team is excited to find out what the future will bring regarding further use of its new AI agents. "We're already live in over 20 countries, and we want to expand further," García says. "Manufacturing is a sector where there's real potential to improve and automate processes, and I think we'll be seeing more AI agents with very industry-specific and vertical capabilities. When you design and scale these vertical-focused AI agents, you really need to bring in domain experts from the customer who truly understand the business, the processes, and the user pain points. Then you bring in an AI platform and trusted partner, like Go Autonomous, and design the experience together."
Both are looking forward to continued strong collaboration. "This whole process has been very transparent, honest, and impactful," García says. "Both companies made a real effort to understand each other's needs. We're looking forward to seeing where the future takes us."
Danfoss is a Danish multinational engineering company and a global leader in energy-efficient solutions that increase machine productivity, reduce emissions, lower energy consumption, and enable electrification. Serving customers in over 100 countries, Danfoss' solutions are used in areas such as refrigeration, automotive, marine, and off- and on-highway equipment.
Industry: Manufacturing
Location: Denmark
Products: Google Kubernetes Engine, BigQuery
About Google Cloud partner- Go Autonomous
Go Autonomous is a Copenhagen-based SaaS company, founded in 2020, leading the emerging category of Autonomous Commerce. Its platform empowers enterprises to digitize and automate their transactional workflows within B2B email – the largest business communication channel – unlocking significant gains in efficiency, productivity, and customer experience. Go Autonomous is driven by a vision to revolutionize B2B commerce with AI-powered agents, providing a seamless, efficient, and scalable platform that simplifies how companies interact, transact, and work. The company is on track to handle approximately €35 billion in annual transactional volume for its customers.
