'Guard the store. Read its pulse.' Why one platform is more honest than two.

The standard pattern in retail technology is two separate vendors. One sells you 'AI loss prevention.' One sells you 'retail analytics.' They sell against each other for camera real estate, even though the cameras themselves are perfectly capable of doing both jobs at once.

Why the industry separates them

The honest answer is margin. Two vendors means two contracts. Two integrations means two professional-services engagements. Two products means two upsell paths. The separation serves the vendors, not the operator.

There's also a historical reason, security cameras and analytics cameras grew up in different industries. The security side came from CCTV vendors. The analytics side came from in-store retail-tech startups. They built separate stacks because their companies were separate.

What changes when one platform does both

Three things matter operationally when the same camera feed drives both detection and analytics.

1. One integration. One installation. One set of cameras to maintain. The IT and ops footprint is smaller and the marginal cost of adding the second product is near-zero.
2. Shared face graph. The faces flagged by the security side are the same faces counted by the analytics side. Whitelisted staff don't trigger alerts but also don't count toward footfall, both correct.
3. Shared audit chain. Everything that happened today, every detection, every count, every approval, is on the same timeline.

The naming

We split the platform into two products with explicit names: Guard for security, Pulse for analytics. They live on the same edge node, read from the same camera feed, share the same data model. From the operator's perspective they're switchable per store, turn one on, turn the other off, or run both. From the architecture's perspective they're one system.

Why the split costs operators money.

When the industry pretends these are separate problems, operators pay twice. They pay for two installations, two contracts, two integrations, and they get two ops teams to manage. The technology doesn't require any of that, the business model does.