AI & Telecom | Dispatch Agentic AI

On this episode of the TeleGeography Explains the Internet podcast, Contrivian's Grant Kirkwood and Tom Daly explain why failover is a failure in itself and how software can combine fiber, mobile, and multiple LEO constellations into one resilient connection. They also cover the quirks of LEO networks and what this means for first responders, energy providers, and enterprises. This conversation covers: Failover vs. active-active connectivity: Kirkwood and Daly argue that failover is a destructive event because users are already affected when it triggers. Monitoring beyond "is the link up?": Daly explains why a simple ICMP ping isn't enough, and what their platform does instead. AI's demands on the network: Why enterprises can no longer just throw bandwidth at the problem and what they must focus on. How LEO has evolved for enterprise: How LEO has moved from a remote-area curiosity to a core part of the network. Contrivian Constellation and multi-constellation bonding: This software orchestrates multiple satellite constellations. Its aim is to create "diversity going up instead of sideways," in place of the old fiber-route diversity that depended on KMZ files and carrier maintenance schedules. Technical challenges of blending access types: Fiber, mobile, and LEO. Where the LEO market is heading: Performance keeps improving, and new inter-satellite and ground-to-ground services are expected to launch starting next year.  Show Notes About Contrivian & New Government Solutions Division   💻 Learn about TeleGeography: https://www2.telegeography.com/ 📊 TeleGeography Resources: https://resources.telegeography.com/ 🎧 About the Podcast: https://www2.telegeography.com/telegeography-explains-the-internet-podcast

On this episode of TeleGeography Explains the Internet, Rohit Jha, co-founder of Transcelestial, explains how his company is using laser links to bring fiber-like speeds to places where digging trenches isn't practical, from storm-damaged Taiwanese villages to chronic outage sites in the U.S. He also lays out the vision behind its planned orbital ring, a network of satellites meant to work like an undersea cable in space and to reshape connectivity for island telcos and AI data centers by the end of the decade. This conversation covers: The "internet distribution problem": How Transcelestial aims to fix under-connected regions that have slow speeds, high costs, and high latency to the cloud. Terrestrial laser links as a fiber alternative: The company's first product, Centauri: a shoebox-sized laser terminal that works as "wireless fiber." Telecom and ISP use cases: Including greenfield deployments that relieve RF congestion on 5G towers with 10x to 100x more bandwidth and brownfield deployments that provide spectrum-free backup when fiber is cut by construction, typhoons, or earthquakes. Enterprise applications: Such as on university campuses, smart factories, mining, and government and defense. Disaster recovery and resilience: How Transcelestial’s laser terminals are used to quickly restore critical connectivity following natural disasters. The orbital ring network: The company’s project to build an "undersea cable in space" by creating a ring around the planet to act as a global transporter network for routing laser-based connectivity. 💻 Learn about TeleGeography: https://www2.telegeography.com/ 📊 TeleGeography Resources: https://resources.telegeography.com/ 🎧 About the Podcast: https://www2.telegeography.com/telegeography-explains-the-internet-podcast

As artificial intelligence transitions from broad hype to practical application, enterprise network managers are asking a critical question: How does AI actually fix my day-to-day workflow?   In this episode of TeleGeography Explains the Internet, host Greg Bryan sat down with Dennis Thankachan, CEO and Co-Founder of Lightyear, to explore how agentic AI is transforming the notoriously painful processes of telecom sourcing, provisioning, and lifecycle management. They discuss: The AI Data Problem: Why standard Large Language Models (LLMs) fail at telecom networking, and how Lightyear uses its proprietary historical data to deliver accurate, context-specific results without exposing customer data. The Shift to Agentic AI: How Lightyear is evolving from a traditional software platform into an "agentic platform" for full telecom life cycle management. The Launch of Layer8: Why Lightyear is creating a new conference specifically tailored to the physical network and enterprise WAN managers. The Future of Network Engineering: How small network teams can leverage AI tools to achieve 10x to 50x productivity improvements as the Wide Area Network (WAN) becomes "cool again." Show Notes Learn more about Lightyear. Register for Layer8 💻 Learn about TeleGeography: https://www2.telegeography.com/ 📊 TeleGeography Resources: https://resources.telegeography.com/ 🎧 About the Podcast: https://www2.telegeography.com/telegeography-explains-the-internet-podcast

Today's guest is Carlie Carson, Research Analyst on the enterprise team at TeleGeography. Carlie and host Greg Bryan consider: How has AI inferring impacted the enterprise network? Is SD-WAN still relevant? Has MPLS disappeared from the enterprise WAN? The pair leverage their own Enterprise Network Trends and Strategy Report to bring actual data to the trends we often talk about on the pod. Carlie and Greg discuss the shifting underlay landscape as MPLS usage drops and Dedicated Internet Access (DIA) takes center stage, why enterprises choose DIA over business broadband, and the real impact (or lack thereof) that AI is having on current enterprise bandwidth demands. They also unpack the reality of satellite connectivity adoption, the evolution of SD-WAN, SASE, and Zero Trust, and why Network as a Service (NAS) is still struggling to gain widespread traction.   Podcast HQ: https://www2.telegeography.com/telegeography-explains-the-internet-podcast TeleGeography Resources: https://resources.telegeography.com/ Our Research: https://www2.telegeography.com/en/our-research

As we look at the massive AI boom sweeping across the globe, what does it actually take from a physical standpoint to keep these models running? Why are submarine cable operators building networks just in case a surge of traffic arrives, and how is local power grid availability dictating exactly where the next major hubs are being built? What role are legacy crypto-mining data centers playing now in AI infrastructure? Today we're joined by a trio of TeleGeography experts to close out our series on the impacts of AI on global telecom. Host Greg Bryan sits down with Alan Mauldin, Patrick Christian, and Jon Hjembo to tie together the network, cloud, and data center layers of this infrastructure puzzle. The group weaves together three TeleGeography research areas that have direct impacts on each other and the expansion of AI services: The Long-Haul Network Buffer: Alan provides a crucial perspective on international bandwidth demand, reminding us that heavy networks cannot move at the speed of software. He explains why providers are securing fiber right now purely as a risk-mitigation strategy against future uncertainty. The Rise of NeoClouds/GPU-as-a-Service: Patrick breaks down how a post-ChatGPT surge overwhelmed hyperscalers, leading to a boom in specialized "NeoCloud" providers—many of which evolved from legacy crypto-mining facilities located far from population centers. Data Center Realities and Power Grids: Jon takes us inside the data center, revealing how a 35% pipeline surge between 2024 and 2025 has collided with intense local regulatory, water, and power bottlenecks in markets ranging from Dublin to Malaysia. We also dive into the fascinating "Pax Silica" geopolitical chip alliance. This special episode is sponsored by Layer8 by Lightyear. This new event will bring together enterprise WAN operators, network engineers, and other IT/infrastructure leaders on November 4 at the Arlo Williamsburg in New York City. Notably, it's free to all enterprise end-users. Learn more and save your spot here.   Podcast HQ: https://www2.telegeography.com/telegeography-explains-the-internet-podcast TeleGeography Resources: https://resources.telegeography.com/ Our Research: https://www2.telegeography.com/en/our-research

The TeleGeography Explains the Internet podcast welcomes Jezzibell Gilmore, General Manager, Service Provider at Kentik. She discusses the profound impact of artificial intelligence on global network infrastructure, and how networks are evolving to handle new technological demands. Key topics covered: The Strain of AI workloads: The explosion of AI and GPU-heavy tasks is exposing the criticality of physical network infrastructure by demanding higher capacity, strict reliability, and ultra-low latency. The rise of network intelligence: As global networks become increasingly complex, the industry is shifting from passive monitoring to "network intelligence," which uses AI to synthesize massive amounts of data and provide actionable insights for operators. Shifting economic dynamics: The traditional expectation of constantly decreasing telecom prices is being upended, as the high demand and limited supply for AI-capable network capacity are actually driving prices up. Physical limitations and the future: Overcoming unchangeable physical constraints—like the speed of light and the distance to new power generation sites—requires extreme network optimization, ultimately pointing toward a future goal of fully autonomous, self-healing networks. Podcast HQ: https://www2.telegeography.com/telegeography-explains-the-internet-podcast TeleGeography Resources: https://resources.telegeography.com/ Our Research: https://www2.telegeography.com/en/our-research

Host Greg Bryan welcomes Phill Lawson-Shanks, Chief Innovation Officer at Aligned Data Centers. With several decades of experience in digital infrastructure, Phill has seen the industry evolve from mainframes and green screens to the massive hyperscale cloud platforms we rely on today. In this episode, we explore: The Shift in Data Center Geography: Why the "center of gravity" for networks is moving toward new power-rich zones and how the rise of AI is rewriting the rules of connectivity. The Engineering Challenges of AI: From floor loading for 6,000-pound racks to the transition from air to liquid cooling, Phill explains the "fungible" design needed for modern AI factories. The Infrastructure Bottleneck: Why AI is ultimately a network-dependent technology and how high-speed optical networking between buildings is pushing against the limits of physics. Sustainability and Power: How the industry is chasing power availability and pioneering new ways to build modular, carbon-traceable infrastructure. Podcast HQ: https://www2.telegeography.com/telegeography-explains-the-internet-podcast TeleGeography Resources: https://resources.telegeography.com/ Our Research: https://www2.telegeography.com/en/our-research

This episode features Hunter Newby, a veteran of the telecom industry with experience spanning from LDDS WorldCom to building carrier hotels and neutral internet exchange points. Newby discusses how AI inferencing is reshaping network infrastructure requirements and why geography still matters in the internet age. Key topics covered: Low-latency inferencing: Why AI applications require a fundamental shift from the old CDN model to distributed, proximity-based network architecture with deterministic routing in sub-millisecond "latency zones" Internet exchange gaps: How many U.S. states lack even a single neutral internet exchange point, forcing local traffic to backhaul hundreds of miles and creating economic development barriers Connected Nation Internet Exchange Points (CNIXP): Newby and Connected Nation are building purpose-built internet exchange points in underserved markets (125+ cities identified) to support enterprise needs and AI workloads, starting with aerospace companies like Airbus and Boeing. Beyond the data center: Why the industry needs to shift focus from power-first mega data centers to network-first interconnection facilities that enable local traffic exchange and support emerging AI use cases Premium routing opportunity: How carriers can move beyond commoditized transit pricing by offering guaranteed low-latency routing to specific zones—the "FedEx model" for data delivery From this episode: Akamai Boosts Inference With ‘Thousands’ of Nvidia Blackwell GPUs Podcast HQ: https://www2.telegeography.com/telegeography-explains-the-internet-podcast TeleGeography Resources: https://resources.telegeography.com/ Our Research: https://www2.telegeography.com/en/our-research

What are the real bottlenecks in AI infrastructure development? How is this infrastructure boom similar to the 90s internet boom? How do we overcome the pricing paradox in telecom transport where demand keeps rising and service prices keep falling? Today on TeleGeography Explains the Internet, we welcome Luis Colasante, Head of Procurement Strategy for Energy & Infrastructure at Colt Technology Services. Luis brings a perspective from the intersection of energy strategy, critical infrastructure, and capital markets. In this episode, we move beyond the "compute bubble" to discuss why physical infrastructure—from subsea cables to the power grid—has become the primary bottleneck for the AI revolution. Luis explains: The Energy-Connectivity Nexus: Why AI data centers require two to three times more power than traditional cloud facilities and how energy availability is now the ultimate gatekeeper for digital expansion. Shifting Investment Cycles: A look at the parallels (and differences) between the late-90s telecom bubble and today’s hyperscaler-led boom. Digital Sovereignty: Why governments are treating subsea cables as strategic national security assets, highlighted by the French government’s recent move with ASN. The Death of the "Toll" Model: Why selling raw bandwidth has become a deflationary commodity business and how the industry is pivoting toward intelligent service layers and "Network as a Service" 2.0. Podcast HQ: https://www2.telegeography.com/telegeography-explains-the-internet-podcast TeleGeography Resources: https://resources.telegeography.com/ Our Research: https://www2.telegeography.com/en/our-research

Host Greg Bryan welcomes Jason Gintert back to the show. As the incoming President of the U.S. Networking User Association, Jason dives into the current state of AIOps in network management. This includes: Low Adoption and Key Challenges: Despite being a popular topic, AIOps adoption remains low due to concerns about trusting AI output (hallucinations) and the indeterminate nature of large language models. Data Integration and Security: The importance of integrating private data using methods like the model context protocol to make AI answers more deterministic. For security, Jason strongly recommends starting with a read-only, zero-trust mindset when exposing data to AI tools. Practical Use Cases: AI is most valuable for root cause analysis of common network problems (e.g., Wi-Fi authentication issues, circuit errors) and providing automated network summaries or predictive analysis. The Human Element: AI is seen as a powerful tool to increase the productivity of network engineers by handling low-level tasks, but it will not replace humans. Engineers remain crucial for exercising judgment and taking responsibility for service-impacting changes. Podcast HQ: https://www2.telegeography.com/telegeography-explains-the-internet-podcast TeleGeography Resources: https://resources.telegeography.com/ Our Research: https://www2.telegeography.com/en/our-research