Andreessen Horowitz is placing two major bets on enterprise AI, but not in the way most people expect. Instead of chasing cutting-edge language models or chip design, a16z is investing in the unglamorous work of automating the tasks that keep companies stuck in spreadsheets and manual processes. On the same day, the firm announced a $38 million Series A for Convey, a platform that lets non-technical operators build AI teammates, and a $15 million Series A for Netris, which automates the networking complexity of AI data centers.

The dual investment reveals something important about where venture capital sees the real value in AI right now: not in the models themselves, but in making those models actually useful for the work that matters to businesses. Global AI spending is projected to reach $2.5 trillion in 2026, yet most of that investment isn't translating into measurable returns in the areas where companies actually operate.

Why Are Operators Spending Hours on Tasks AI Should Handle?

At most companies, capable people spend two to three hours every single day pulling data across spreadsheets, ingesting orders, and running recurring reports by hand. These are the kinds of tasks that were never supposed to be done manually in the first place, yet they persist because the software that could automate them either doesn't exist or requires months of engineering work to implement.

Convey's approach is to let the operators closest to the problem build their own AI solutions without writing code. A non-technical user can onboard a fully functional AI teammate in about three hours, not weeks or months. The platform has already completed over one million hours of automated work for companies including NBCUniversal, Samsara, TelevisaUnivision, Unity, Faire, and ChargePoint.

The results speak for themselves. At Savoya, the platform delivered a 40 percent increase in EBITDA year-over-year, with the company on track to reclaim 10,000 hours of work in 2026. A large streaming service freed up 23,000 hours annually from reporting and advertising operations workflows, time that teams now spend with customers instead of staring at dashboards.

"The ability to think of an idea and create a solution without convincing an engineer or product manager and doing it in a day instead of weeks is amazing," said Jeremy Varner, Senior Vice President of Programmatic Operations at TelevisaUnivision.

Jeremy Varner, Senior Vice President of Programmatic Operations, TelevisaUnivision

How to Understand the Enterprise AI Opportunity

• The Adoption Problem: Enterprise AI's defining failure mode right now is adoption without impact. Usage keeps climbing, but return on investment never materializes. Convey's approach flips this by having AI teammates own the outcome on an organization's highest-impact operational work, making the return measurable from day one.
• The Speed Advantage: Traditional enterprise software implementations take months or years. Convey's no-code approach lets operators deploy AI teammates in hours, with organizations seeing transformative impact within one to two months. This speed matters because every week a company waits is a week of lost productivity.
• The Governance Challenge: Enterprises can't just let anyone build AI solutions. Convey partners with IT teams to configure each teammate's identity and permissions, connecting to critical company systems including legacy tools without APIs, ensuring security and governance while empowering operators.

Rohan Chopra, Convey's founder and CEO, spent eight years at DoorDash, where he watched capable operators spend their days on manual work that software hadn't caught up to. That pattern repeated in nearly every enterprise he encountered afterward. Convey is built to close that gap.

"The companies winning right now have figured out how to remove the operational drag that keeps their teams from doing strategic work. They are scaling by onboarding a digital workforce that creates real, measurable impact," said Rohan Chopra, Founder and CEO of Convey.

Rohan Chopra, Founder and CEO, Convey

Joe Schmidt, a partner at Andreessen Horowitz, is joining Convey's board. He emphasized that the firm sees this as a fundamental shift in how enterprises adopt AI. "Enterprise AI's defining failure mode right now is adoption without impact. Convey teammates own the outcome on an organization's highest-impact operational work, making the return measurable from the start," Schmidt noted.

What About the Infrastructure Layer? Netris Solves the Networking Bottleneck

While Convey tackles the application layer, Netris addresses a different but equally critical problem: the networking complexity of AI data centers. As companies race to build AI infrastructure, they're discovering that spinning up a data center isn't just about securing GPUs and storage. Getting everything configured, running, and able to serve multiple customers can take months.

Netris provides software that automates network setup, configuration, and operations for AI clusters. The platform abstracts the complexity of managing multiple network fabrics at once, Ethernet, InfiniBand, and specialized AI-focused fabrics, each with its own control plane. When an operator adds, resizes, or removes a tenant, all those fabrics have to be reconfigured in concert, and one misconfiguration can take a cluster down or leak one customer's data into another's.

The company calls its approach NAAM, which stands for Network Automation, Abstraction, and Multi-Tenancy. It's become the platform that neoclouds standardize on as this category takes shape. Netris is already live at more than 35 GPU clusters around the world, operating roughly one million GPUs in total, deployed by companies including Lightning AI, Foxconn, Visionbay, Hewlett Packard Enterprise, TensorWave, Telus, and others.

In the last 12 months, Netris experienced 800 percent annual recurring revenue growth, driven by strong demand from AI cloud providers looking for better software abstraction tools. The company has a strong partnership with Nvidia, which recommended Netris to customers two years ago after being impressed by a demo of the technology.

"As a GPU cluster operator, you need to make configuration changes to every link, every day. For AI, software is not okay, because the amount of traffic is so high, everything must be hardware accelerated. This is what we do, and this is what we've been doing for eight years," said Alex Saroyan, CEO of Netris.

Alex Saroyan, CEO, Netris

Guido Appenzeller, a general partner at Andreessen Horowitz, led the Netris round and is joining the company's board. Appenzeller noted that every era of computing has needed a new networking foundation. "First for the virtualized data center, then for the cloud, and now for AI. GPU clusters run across many fabrics at once, and legacy automation was never built for that," he explained.

Appenzeller

The a16z team backing Netris includes Martin Casado and Raghu Raghuram, both of whom helped build the last generation of data center networking. Casado founded Nicira, the startup that pioneered software-defined networking and was acquired by VMware in 2012. Raghuram rose to become VMware's CEO. Their involvement signals that a16z sees the AI infrastructure challenge as equivalent to the data center revolution they helped lead.

What Does This Mean for the Broader AI Market?

These two investments reveal a strategic shift at one of Silicon Valley's most influential firms. Rather than betting on the next breakthrough AI model or the next generation of chips, a16z is betting on the operational and infrastructure layers that make AI actually work in the real world. The firm is essentially saying that the bottleneck isn't intelligence anymore; it's implementation.

For Convey, the $38 million Series A includes continued participation from Khosla Ventures and Pear VC. For Netris, the $15 million round is supported by a16z's top infrastructure and networking partners. Both investments suggest that a16z sees massive market opportunities in solving the practical problems that enterprises and infrastructure operators face every day.

The timing matters. With $800 billion expected to be spent on AI infrastructure this year and trillions to be deployed over the next few years, the companies that can help others deploy, configure, and operate that infrastructure at scale will capture enormous value. Convey and Netris are betting that they can be those companies, and a16z is betting alongside them.