Netris just closed a $15 million Series A led by Andreessen Horowitz to solve one of AI infrastructure's thorniest problems - getting new data centers online fast enough to meet exploding demand. The startup's network automation software promises to slash deployment times for so-called neoclouds, the wave of specialized AI compute providers racing to compete with hyperscalers. With AI training clusters now requiring thousands of GPUs and ultra-low-latency networking, Netris is betting the infrastructure layer needs a complete rethink.
Netris just secured $15 million in Series A funding from Andreessen Horowitz, betting that the AI infrastructure gold rush needs better shovels. The timing couldn't be sharper - as companies scramble to build GPU-packed data centers to power AI workloads, the networking layer has become a critical bottleneck.
The Santa Clara-based startup provides software that runs directly on network switches, turning what's typically a months-long manual configuration process into something closer to plug-and-play. For neocloud operators - the emerging class of specialized AI compute providers trying to challenge the hyperscaler oligopoly - that speed difference is existential. Every week of delayed deployment is revenue left on the table in a market where GPU capacity commands premium pricing.
"We're seeing operators who want to light up new AI clusters in weeks, not quarters," the funding announcement suggests, though specific customer timelines weren't disclosed. The platform abstracts away the complexity of managing network fabrics, load balancers, and routing protocols that traditionally require specialized engineering teams to configure and maintain.
The neocloud category has exploded over the past 18 months as AI training and inference workloads outpace what traditional cloud providers can deliver. Companies like CoreWeave, Lambda Labs, and a growing cohort of regional players are building purpose-built infrastructure optimized for AI - dense GPU clusters, high-bandwidth interconnects, and liquid cooling systems that look nothing like conventional data centers.
But hardware is only half the equation. Network configuration for AI clusters presents unique challenges - training runs for large language models require near-perfect synchronization across thousands of GPUs, where even microseconds of latency can torpedo performance. Traditional networking tools weren't built for this scale or these requirements, creating an opening for specialized solutions.
That's where Netris fits. The company's software essentially provides a cloud-like control plane for physical network infrastructure, letting operators manage switching fabric through APIs and automation rather than command-line interfaces and manual configs. Think of it as Kubernetes for the network layer - abstracting complexity while giving operators the flexibility to optimize for AI workloads.
Andreessen Horowitz's involvement signals continued appetite for infrastructure plays in the AI stack, even as some investors grow cautious about application-layer startups facing uncertain unit economics. The firm has been systematically backing companies across the AI infrastructure landscape, from chip design to orchestration software, betting that whoever owns the plumbing will capture outsized value as the market matures.
The $15 million round positions Netris to scale its engineering team and expand beyond early adopter customers. With neoclouds projected to capture a growing share of AI compute spending - some analysts estimate the category could reach $20 billion in annual revenue within three years - the race is on to become the default networking layer for this new generation of infrastructure.
Competition in the space is heating up. Traditional networking vendors like Cisco and Arista are adapting their portfolios for AI workloads, while startups are attacking specific pain points across the stack. But Netris is betting its neocloud focus and software-first approach will resonate with operators who need speed and simplicity over enterprise feature bloat.
The funding also arrives as broader questions swirl around AI infrastructure buildout sustainability. With Nvidia GPUs still in short supply and power constraints limiting data center expansion in key markets, investors are scrutinizing whether the neocloud wave represents durable disruption or a temporary arbitrage opportunity. Netris's thesis - that specialized infrastructure operators will need specialized tooling - hinges on the category achieving escape velocity.
What happens next depends partly on factors outside Netris's control. If neoclouds successfully carve out sustainable niches against hyperscaler competition, demand for deployment acceleration tools should follow. But if cost pressures or market consolidation winnow the field, the addressable market could shrink faster than expected. For now, a16z is placing a bet that the infrastructure layer still has room for innovation - and that getting AI clusters online faster is worth paying for.
Netris's Series A captures a broader shift in how AI infrastructure gets built - speed and specialization increasingly trump the one-size-fits-all approach of traditional cloud providers. Whether the neocloud category achieves the scale investors are banking on remains to be seen, but the underlying demand signal is clear: companies need AI compute capacity yesterday, and whoever can deliver it faster stands to win. For Netris, the challenge now is proving its automation platform can scale alongside its customers' ambitions while fending off competition from both established networking giants and fellow infrastructure startups chasing the same opportunity.
