Milestone Systems Innovation & Technology Culture

How does innovation show up in your company culture?

The engineers on Milestone’s Arcules team are motivated to deliver technology that makes our users’ jobs easier. Our remit is incredibly broad — video, audio, devices, networks, data, AI, UX and cloud-based SaaS. For savvy developers, the opportunities are boundless. As an engineering manager with decades of hands-on experience, I’m serious about our core responsibility: protecting our engineers' ability to innovate and ship. 

For us, innovation takes many forms. An engineer-led working prototype led to an entirely new UX experience. A new service implementation drove application-wide removal of tech debt. In fact, one of our major product differentiators was created by a developer who simply asked if there wasn’t a simpler way. Engineering voice and leadership are valued here; a functional demo is worth a thousand words. 

We run lean with a few excellent product managers and zero project managers. This requires staff to manage, communicate, and collaborate, and we do it all remotely across North and South America, and occasionally European, time zones. This isn’t the right environment for everyone, but for us on the Arcules team, we wouldn’t have it any other way.

What’s one recent innovation that improved user or employee experience?

Just this week, we released an AI-assisted forensic search capability that transforms our users’ ability to find exactly what they are looking for via free-text search of incoming video streams. This shifts our users’ experience from passive video review to active and dynamic search when someone has a safety or security concern. 

The challenges are plentiful: interpreting video accurately and quickly; matching interpretations to each frame at scale; offering query capabilities over wide time domains; building a UX that lets users narrow and widen a search intuitively; and scaling inference in real time while managing cost. 

Our commitment to responsible AI is non-negotiable. Forensic search users always see the source image and video alongside any AI finding to confirm or deny it themselves. We moderate use through automated and manual review of search terms and via terms of service. We build with a focus on ethical use and ensure our product meets European AI Act standards. Our stretch goal is a fully dynamic user interface that learns the behavior of individual users and anticipates their needs. At Milestone, we think of our job as helping the world to see.

How do you balance experimentation with stability?

Our informal motto is “Move fast and make things.” Our worldwide customers require a stable experience. As an engineering team that is responsible for our own quality assurance, we know the buck stops with us. For every effort, we know we must ensure stability. We believe in outcomes over output. 

We use feature flags. When appropriate, we apply version-based rate limiting to effect a gradual shift and watch for errors; sometimes we use the front end to communicate contextual information to the back end to ease a transition. Underlying all these techniques is our commitment to a flow-oriented model, which focuses on finishing work in progress before starting anything new. We build large-scale systems and assemble them piece by piece in production, allowing us to experiment, improve, and harden code before users ever see it. 

The most exciting experimentation approach involves partnering with our customers. We spin up what we call our “Flywheel” effort, meeting with customers for a show and tell of experiments and refinements, so that when a concept really resonates, we can create and ship a feature that is both fresh and mature.