When the Network Fails, Everything Else Does Too

Artificial intelligence is reshaping how businesses operate, but the networks behind it determine whether that transformation delivers or stalls. In this episode of Go Beyond the Connection, Joel Mulkey, Founder and Board Member at Bigleaf Networks, makes the case for resilient adaptive networks for AI as the foundational infrastructure every organization now depends on. His argument is direct: when AI tools go offline, so do the people relying on them.

Mulkey traces the origins of Bigleaf to a real outage at a church service he was running sound for, when a Comcast failure knocked out every system connected to the cloud. That experience crystallized a vision: build internet redundancy that works for everyone, not just those with Cisco-certified engineers on staff. The result was a cloud-native tunnel architecture that monitors every connected circuit and reroutes traffic automatically, keeping sessions intact and users uninterrupted.

How Adaptive Monitoring Changes the Resilience Equation

Traditional failover tools were built for lab environments, not production. Bigleaf’s approach takes a fundamentally different path: a device at each customer location creates tunnels to cloud-based gateways, and the system measures latency, packet loss, and jitter on every circuit ten times per second. When a path degrades, traffic moves without changing the user’s IP address. Calls continue. Transactions complete. AI workflows keep running.

Our software can detect problems along that whole internet path end-to-end and move the traffic around in real time with the user not even noticing because their IP addressing doesn’t change. -Joel Mulkey

This design also reframes how organizations think about wireless connectivity. Mulkey challenges the assumption that wireless is inherently second-tier, pointing out that with adaptive monitoring, 5G and low-earth-orbit satellite connections can match or exceed the reliability of cable or DSL. For multi-location businesses operating in areas where fiber is not feasible, that shift opens options that simply were not available before.

AI Makes Every Downtime Decision Costlier

As organizations integrate AI tools into daily workflows, connectivity becomes load-bearing infrastructure in a new way. When a user depends on an AI assistant to answer a question that used to take five minutes of documentation review, losing the connection does not just slow them down. It stops them. For companies running private AI clusters or training proprietary models, the stakes are even higher: a dropped session can mean hours of compute time lost.

Mulkey frames resilient adaptive networks for AI not as a premium feature but as a baseline requirement, and he offers a practical challenge to every IT and business leader: calculate the actual cost of five minutes of downtime. One hour. Five hours. Once those numbers are visible, the business case for proactive resilience becomes straightforward.

EPISODE HIGHLIGHTS

• How cloud-based tunnel architecture routes around failures without dropping active sessions.
• Why wireless connectivity, with intelligent monitoring, now belongs in the enterprise circuit mix.
• The two AI use cases that make network uptime more mission-critical than ever: cloud tools and private clusters.
• How behavior-based traffic prioritization protects voice, video, and SaaS tools without manual rule management.
• The people-first values that have shaped Bigleaf from its founding through its growth.

Network resilience is no longer a back-office IT concern. It is a strategic business requirement. Listen to or watch this episode of Go Beyond the Connection to hear how Joel Mulkey built Bigleaf around the conviction that every organization, regardless of size or technical sophistication, deserves connectivity that works when it matters most. Subscribe to the podcast for more conversations with leaders shaping the future of enterprise connectivity.