Boulder Ai

"We beat INTEL by weeks for our PATENT"
Boulder AI lives at the convergence of two megatrends. Deep learning made visual AI accurate while edge
computing made it scalable.
With a general-purpose, edge visual intelligence platform, Boulder AI can simulate human insight and
decision making on a device. For the customer, this delivers insights that weren’t possible before—superior
information for operations teams, better insights for planning, and automated real time decisions.
For Boulder AI, it signals an endless opportunity to sell SaaS AI applications on standard devices in diverse
environments, providing insights that weren’t possible with forensic focused Physical Security systems or
siloed IoT devices.
Boulder AI was founded three years ago by a team of visual AI engineers, hardware designers, and the
business team behind Avigilon.

The Challenge
Today, information about a physical environment typically comes from people who are observing live events
or reviewing activities on video systems. Because people can’t observe places all the time, operators of
cities, worksites, retail stores, casinos and almost every other setting make important decisions based on
anecdotal information.
In the future, nobody should stake their lives, dollars, or time on incomplete information. This was the
objective behind the IoT and Physical Security sectors that was never achieved with legacy technologies.
With the accuracy and adaptability of deep learning, visual sensors are now capable of simulating human
observation.
The Future
Boulder AI puts human insights and decision making on a visual sensor, allowing customers to augment
their workforce at the price of an IP camera.
Our product includes general purpose, edge computing visual sensors, a SaaS deep learning application
store that tailors visual sensing to the customers’ environment and insight needs, and a cloud platform for
data visualization and remote system management. Collectively, this visual intelligence platform delivers
insights that weren’t possible before—informing operations teams, improving planning, and allowing
customers to automate routine decisions with simulated human oversight.
Leadership
Boulder AI’s business team previously defined the physical security market. Bryan Schmode, CEO, has a 20
year track record of success in the physical security industry. He previously served as COO of Avigilon, a 
leading physical security company that went public in 2011 and drove it from $0 in revenue to over $400
million in annual revenue. Avigilon was acquired by Motorola Solutions in 2018 for $1 billion.
Darren Odom, CTO, founded Boulder AI in 2017 and invented the DNN Camera technology. Darren holds a
BS EE from the Colorado School of Mines, and has a decade of experience in computer vision. Darren built
an engineering team composed of experienced computer scientists, data scientists, software engineers, and
electrical engineers.
Boulder AI’s sales organization is headed by Pat Reilly, who previously held VP of Sales roles for six
companies that exited under his tenure. He drove the expansion and repeatability of sales at the following
companies sold to Safenet (Cylink), SAIC (Cloudshield), Cisco (Broadware), Tyco (Proximex), Watchguard
(Hexis) and Oracle (Talari).
Competition
The majority of incumbent OEM’s sell visual analytics that centrally process IP camera streams. These
remote, server-based visual analytics systems are expensive due to server, networking, installation and
maintenance costs. They have privacy, security flaws and latency which rules out local collaboration and
automation.
Edge computing is disrupting visual analytics because processing at the image source yields the resolution
that deep learning needs, without costs from a supporting server, bandwidth, and storage infrastructure.
Some OEM’s are beginning to build edge computing visual sensors, but the existing market is heavily
focused on low cost devices optimized for a single application (ASICs). As a result, their business model
remains entrenched in the IoT and Physical Security practice of selling hardware.
In contrast, Boulder AI shifts the sale from hardware to more software and services. Our customers can
easily configure any of their edge devices to run new applications that are available in the App Store and
easily deploy them from the central Boulder AI Dashboard. We provide the first general purpose visual
computing platform, which delivers AI flexibility for any environment with our application store, high accuracy
at low cost, reduced operational cost with over the air updates and retraining, and privacy by design.
Market Size
Boulder AI’s visual intelligence platform marries the benefits of IoT and Physical Security. We are changing
customer expectations with our systems by bundled sensing, AI and security with their device purchase.
This $84 billion forensics focused Physical Security and $164 billion IoT markets do not represent the full
extent of Boulder AI’s opportunity.
Since Boulder AI devices run multiple applications and can be updated in the field, each device represents
ongoing opportunity to upsell subscription software applications that provide users with services that didn’t
exist before. Boulder AI anticipates expanding the visual computing market by reaping the benefit gained by
enhancing worksite safety (to decrease $60 billion in annual worksite injuries), reducing traffic hazards (the
source of $871 billion in annual losses), minimizing casino losses from insider theft (>$1M/year per casino),
and helping organizations manage the magnitude and social distancing of employees and customers after
Covid-19 SIP orders are relaxed saving lives and giving people the confidence to resume some semblance
of a normal life/work experience.
Marketing Plan
As a network-fueled business, Boulder AI’s success hinges upon rapid adoption. We are driving adoption by
differentiating our solution for channel partners who own relationships across our target verticals and
geographies. Currently, we have installations that deal with Smart Cities, the environment, manufacturing,
Oil and Gas and we are now seeing traction in the Gaming market.
When channels and integrators sell conventional IoT and Physical Security solutions, they only make a
margin on hardware, software licensees, and installation/maintenance. In contrast, Boulder AI provides
additional revenue streams to our partners.
Boulder AI designates a portion of the Recurring Monthly Revenue (RMR) to the channel partner, yielding
predictability and enterprise value. Because our solution supports multiple applications, our partners can
upsell SaaS products to existing customers and increase annually recurring revenues. Collectively, the
channel partner benefits from predictable and increasing SaaS revenues, maintaining touch with their
customers’ needs and use cases, and good margins from system sales and integration.
Traction
Boulder AI has two generally available devices, the DNN Node and DNN Camera. The DNN Node ingests
streams from legacy IP cameras, while the DNN camera integrates a high-resolution sensor and
general-purpose supercomputer. We have sold hundreds of these devices across the globe to dozens of
channel partners that are working on POC’s with target customers. Many of these efforts are now
transitioning into initial deployment efforts.
In Q1 2020, we launched Boulder AI’s cloud platform. And, currently, we have multiple applications
developed by Boulder AI and third parties available in the App Store.
Boulder AI has several high profile, paying customers with significant deployments and paths to scale. In
particular, Boulder AI is partnered with the City and County of Denver, and has significant recurring revenue
from installations across the City.