"The importance of satellite networks is growing, but they will not replace existing networks due to physical limits. They will play an important role as a complement."
Ishwar Parulkar, chief technologist for communications at Amazon Web Services (AWS), said in a video interview with ChosunBiz on the 24th that the non-terrestrial network (NTN) is evolving rapidly. Parulkar added that the opportunity area where carriers can generate revenue going forward is personalized agentic AI that uses data.
Parulkar joined AWS in 2016, convinced 10 years ago that the cloud could fundamentally transform the telecom industry. Before AWS, Parulkar worked at Cisco as a lead architect in the communications business unit, developing technologies including telecom routers, packet cores, small cells, orchestration, and automation, and also worked at Sun Microsystems and Apple.
Parulkar's role as chief technologist is to set AWS's technology strategy. He works with product teams to define new services and features tailored to communications and partners with major customers to establish cloud strategies and visions. The following is a Q&A.
―How will 6G redefine connectivity?
"When 5G (fifth-generation mobile communications) was being developed, the cloud was something of a secondary consideration. But in 6G (sixth-generation mobile communications), cloud-native means building the architectural foundation in a way that makes management, updates, upgrades, operations, and sharing in multi-tenant environments easier. 6G will go beyond cloud-native and become AI-native. AI will be more deeply integrated into network optimization and management, and the network itself will be designed with AI applications in mind. Physical AI will also be an important part. The physical world—such as sensors, robots, and autonomous vehicles—requires very low latency, and there is also a digital world of digital twins that must interoperate with this physical world. The network will be the key foundation that connects this digital world and the physical world. Also, in 5G, non-terrestrial networks were a separate track, but in 6G they will be integrated much more tightly. In short, cloud-native, AI-native, support for physical AI, and integration of terrestrial and non-terrestrial (NTN) networks are the key dimensions that will define 6G."
―Communications are now expanding beyond terrestrial networks to space-based connectivity.
"The non-terrestrial network space is evolving rapidly. Amazon also began its satellite internet business in 2019, building the Kuiper network and starting this journey. In 6G, discussions are moving toward integrating non-terrestrial and terrestrial networks. In 5G, terrestrial and non-terrestrial networks were treated as separate tracks, but now they are being considered together. While the importance of satellite networks is clearly increasing, I do not expect them to replace existing networks; they will play an important role as a complement. Non-terrestrial networks face physical limits on available bandwidth and spectrum and cannot match the bandwidth provided by buried fiber. Just as 5G saw convergence of wireline and wireless—common network functions across fixed and mobile—I expect a similar trend of integrating terrestrial and non-terrestrial networks in 6G."
―How should carriers generate revenue going forward?
"One of the key means is agentic AI. Agentic AI is being used not only to boost productivity and cut expenses but also to build new products and enable things that were previously impossible. In telecom, it is possible to understand the customer experience more deeply, analyze the network, and combine these insights to upsell and cross-sell beyond simple bundles of connectivity services into more personalized services. Personalization is a major opportunity. Just as media and entertainment and healthcare have achieved strong results with personalized recommendations, carriers have significant opportunities to better understand customers and personalize services and products to increase revenue. Another large but underused asset carriers have is data. There is mobility information, location data, and user authentication data for financial transactions. Carriers have an opportunity to create new revenue streams through data that agentic AI applications can use. Agentic AI is one of the catalysts that makes this path more feasible."
―Korea and the global telecom industry are focused on domestic markets.
"Carriers are interested in expanding overseas, but they generally tend to stay in their home markets. In Europe, there are cases where a large parent company has multiple operating companies (opcos) across the continent and expands its business-to-business (B2B) operations in other markets by leveraging common assets, but it is rare for a carrier to enter a completely new market. SK Telecom formed the Global Telco AI Alliance, an AI consortium, and agreed to jointly develop a "Telco AI Platform" based on core AI capabilities, which can be seen as expanding the market overseas. The Telco AI Platform includes jointly building a core large language model (LLM) to develop AI agents for each carrier in the future."
―What businesses is AWS currently focusing on?
"AWS is recently building and modernizing carriers' IT infrastructure and networks in a fundamentally different way. From infrastructure to tools for building and operating agentic systems to ready-to-use applications, AWS has the full stack for agentic AI, enabling carriers to operate networks in an agentic manner. In particular, we are focusing on network automation and network integration using agentic AI. Agentic AI is driving major change. Alongside this, AWS helps customers migrate to the cloud using the latest infrastructure, purpose-built databases, and modern data architectures, reducing operating expenses and making system upgrades and maintenance much easier."
―What are representative customer cases?
"A representative case of creating a revenue model through network modernization is KT. KT used AWS to modernize its enterprise messaging workflow based on Generative AI. As part of that, it introduced an intelligent AI customer service (CS) assistant capable of handling technical inquiries and service-related questions into CPaaS-UMS, its B2B communications platform. CPaaS-UMS is a concept that integrates, sends, and manages statistics for multiple communication channels through a single API (unified messaging system). KT built the assistant on Amazon Bedrock, AWS's Generative AI app development platform, and Amazon Nova Pro, a high-performance multimodal model. Working with the AWS Generative AI Innovation Center and the AWS solutions architect team, KT created a Q&A knowledge base with more than 580 items. KT grew cloud-based messaging sales from $1.4 million in 2024 to $5.5 million in 2025 and, in collaboration with AWS, is targeting $15 million in 2026."