logo

About Us

Edge Node AI

Inference at the speed of proximity. We're building a distributed data center network across the United States.

Contact uscontact@edge-node.ai

Who We Are

Your models. Your data. Our infrastructure.

Edge Node AI is an infrastructure company deploying inference capacity in U.S. metropolitan markets. Rather than concentrating compute in a few large campuses, we place it near the users and applications that depend on it.

We recognized early that as AI moves from research labs into real-time production environments, the bottleneck shifts from raw compute to latency.

What We Do

Token-as-a-Service. Inference, delivered.

At the core of every Edge Node AI site is our inference token-as-a-service platform — a fully managed software layer that transforms raw GPU capacity into on-demand, metered inference tokens. Companies connect once and get instant access to ultra-low-latency model inference without managing hardware, orchestration, or scaling.

  • Single API endpoint
  • Automatic routing
  • Pay per token
  • Ultra-low latency

Custom Deployment

Dedicated environments for private workloads

For organizations with proprietary models or sensitive data requirements, Edge Node AI offers custom dedicated environments for hosting distillation and fine-tuned models — whether you're distilling a frontier model down to a purpose-built variant or running inference against private datasets.

Isolated compute

A dedicated environment provisioned for your workloads alone.

Reserved hardware

Distilled and fine-tuned models run on hardware exclusively yours.

Fully managed

High-performance infrastructure without building or running it yourself.

End-to-end support

From infrastructure planning and connectivity to ongoing operational support.

Our Approach

How we think about infrastructure

01

Proximity Over Scale

We deploy many smaller, strategically placed facilities rather than a few massive ones. This ensures that no customer is ever far from an inference endpoint, keeping round-trip latency as low as physically possible.

02

Inference-First Design

Our sites are configured for inference rather than training: sustained token generation, not burst-mode model training. We select and equip facilities for the power, cooling and network characteristics inference workloads actually require.

03

Fully Managed Simplicity

From hardware provisioning to model deployment and scaling, our platform handles it all. Customers interact with a single, clean API and never think about the infrastructure underneath.

Leadership

The team behind the network

Our leadership team brings 20+ years of experience designing, deploying, and operating resilient data center environments.

Portrait of Mallik Panda

Mallik Panda

Founder, CEO & CTO

Built martech for the first iPhone launch. Now building the compute layer for AI inference.

Mallik founded iTeam straight out of business school in 2005 and, within two years, was part of the team behind the first iPhone launch at Apple — building the martech platform that carried it worldwide. Twenty years of enterprise work followed, for Hilton, Bank of America, NYSE, Kohl's, and others. When IoT started pushing workloads out of the cloud, he founded Edge Node to put compute where the data actually lives.

Portrait of Narendra Manney

Narendra Manney

President & Promoter

nmanney@edge-node.ai

Two decades in enterprise technology, two more in multi-site infrastructure. Edge needs both.

Narendra has had two careers, and edge infrastructure runs on both. The first was 20+ years in enterprise technology, delivering ERP and RFID implementations across the US, UK, Europe, Singapore, Malaysia, and India. The second was acquiring and operating a large network of fuel and convenience retail locations — a business that lives or dies on real estate, capital discipline, and running hundreds of sites well. At Edge Node he leads growth and capital formation across the 2GW core and the tower network beyond it.

Portrait of Karim Naguib

Karim Naguib

Chief Operating Officer

Spent 27 years building a data center business with his own capital at risk — then sold it.

Karim founded a colocation and managed services company in 1996, ran it for 27 years, and exited in 2023 — construction, customers, operations, and the sale, all under his own ownership. Before and alongside that came senior infrastructure roles at Fujitsu, Microsoft, and Dell. At Edge Node he runs operations and the Dallas facility, and decides which markets we enter and how conventional colocation gets rebuilt to carry AI workloads. B.S. in MIS, UT Dallas. English and Arabic.

Timeline

From incorporation to operating capacity

The build so far, in the order it happened.

  1. Edge Node Inc. incorporated in Texas; team in place

  2. Dallas data center site secured

  3. Colocation deployments begin

  4. 32-node cluster ordered

    Enterprise client engagement signed

Facilities

Inside the network

The sites where the compute actually lives.

  • Edge Node AI facility
  • Edge Node AI facility
  • Edge Node AI facility

Ready to move inference to the edge?

Get in touch to learn how Edge Node AI can bring ultra-low-latency AI to your applications.