As the world continues its sometimes uneasy but expectant shift towards automation and artificial intelligence (AI), organizations across a multitude of verticals are assessing the seemingly endless opportunities.
Tupl, short for ‘tuple,’ a technical term originating from the world of math and programming, is an Intelligent Process Automation (IPA) company that found its start in the telco space a decade ago, before most of the AI buzz began.
It has already been delivering impressive results, according to Rafa Ballesteros, Head of Business & Technology Americas, Tupl. The data-driven company, focused on “disruptive innovation” says it is helping telcos reduce manual labor by 90 percent, increase speeds by up to 100x and accuracy by up to 4x, when compared to existing, manual engineering processes.
Intelligent Process Automation in Telecom
Celebrating its 10th anniversary, Tupl has been leveraging AI to streamline processes that once required manual human intervention. The company with about 70 employees worldwide, is anything but limited in its reach, according to Ballesteros. “We have hubs in Japan, Dallas, and Bellvue, Washington,” he explained, noting that their main R&D office is in Malaga, Spain.
At the heart of Tupl’s mission is IPA, which focuses on solving complex, repetitive tasks within telecom networks. “We handle all types of solutions. If you have a human making repeated decisions by looking at multiple sources of data, that’s where we come in,” Ballesteros said. Sample solutions include RF Shaping, a field-proven AI reinforcement learning agent to reduce interference while maintaining coverage and the Open RAN (ORAN) toolkit which leverages Machine Learning Operations (MLOps) and automation.
The telecom industry is inundated with data, from performance management to coverage maps. Tupl’s AI-driven platform is also designed to identify, analyze, and resolve root causes of network issues. Whether it’s managing network operations or providing technical customer care, Tupl is bridging the gap between manual oversight and efficient automated solutions. “We are able to detect problems before customers even complain,” said Ballesteros, “we use a proactive approach to network issues, rather than the traditional reactive method.”
Tupl’s machine learning models allow it to determine root causes for customer problems and provide quick resolutions, without a dispatch, according to Ballesteros. This includes scenarios where a customer’s service is misprovisioned or when a network performance issue occurs.
Partnering with Industry Giants
Tupl has a long-running relationship with major network operators including a Tier 1 Mobile Network Operator (MNO). “We were the first to deploy a distributed system at a Tier 1 MNO in the U.S.”, said Ballesteros. Over the years, Tupl has developed and refined a suite of tools for a them, including the AI Care Now, a first-call resolution (FCR) enabling care agents to provide a simple response to customers for a complex network issue. “We can root cause customer issues in under 30 seconds,” Ballesteros said. At the Tier 1 US MNO, Tupl’s platform monitors everything from provisioning to porting customers in and out, which alone has saved this MNO an estimated $40 million in the last six months.
Tools include network optimization, technical customer care, and proactive management of power-saving features across all cell sites in the network. With the advent of more energy-efficient equipment in telecom, Tupl’s platform can dynamically manage these new features, ensuring cost savings and improved efficiency, says Ballesteros. “We orchestrate the power-saving features that vendors are developing,” he added, emphasizing Tupl’s ability to bridge multiple vendor systems within a telecom operator’s network.
As the market for AI automation solutions grows, Ballesteros acknowledged the presence of other players, but reiterated that Tupl’s deep industry knowledge and simplified platform give it a unique, competitive edge. “I think that one of our assets is that deep knowledge of telecommunications, sometimes to jumpstart that knowledge is hard,” Ballesteros noted.
Partnering with Tupl, customers can customize solutions. Telecom operators are able to tweak key performance indicators (KPIs) and build machine learning models with flexibility. Tupl’s platform is designed for “scalability and fast development,” said Ballesteros. “Once a customer becomes familiar with the system, the possibilities are nearly endless.”
Looking Ahead: AI Trends in Telecom
As Tupl continues to evolve, Ballesteros highlighted its foray into Large Language Models (LLMs), a trend that is becoming increasingly important in AI development. “We are proud, traditional AI people,” he said, emphasizing that while LLMs are gaining traction for certain applications, traditional AI models still hold significant value in ensuring accuracy and reliability, particularly in telecom.
Accuracy remains critical for telecom operators, where operational safety and precision are non-negotiable. “When you need to prioritize accuracy, you need to go with safe, proven systems,” Ballesteros concluded.
Tupl’s journey in the AI automation space is a testament to the transformative power of technology in telecom. As Ballesteros explained, “It is a journey,” and one that will continue to evolve, but Tupl stands ready to pave the way.
For more information, visit https://www.tupl.com/solutions/telco/.
Reader Interactions