Hyperspace is making stores think and act like websites

Retail tech has spent years trying to make shops behave like websites. Online, every click, hover, and abandoned basket is measured and optimised until intent becomes revenue. In a supermarket or mall, that precision often vanishes: managers rely on instinct, a few CCTV feeds and yesterday’s numbers to guess what is happening now.

The offline blind spot in a data-driven retail world

Ulisse Ltd, led by CEO Luca Nestola, argues that blindness is expensive in Southeast Asia, where competition is fierce, and service expectations are rising. Its platform, Hyperspace, is built around what he calls “Physical AI”: software that analyses how crowds move and prompts staff to act before friction turns into lost sales at the shelf and checkout.

Also Read: Why retailers must think like tech companies to thrive in a data-driven economy

Nestola’s starting point is a mismatch between the digital and physical worlds. E-commerce tracks the customer journey end to end and adjusts constantly. Physical retail, despite accounting for about 85 per cent of sales, still runs with limited visibility.

Traditional analytics, often based on cameras or manual counts, tends to answer only what happened: how many people entered, where they went, and when footfall spiked. By the time reports land, the moment has passed.

Hyperspace is meant to behave more like a live control system, matching demand with capacity as conditions shift. In-store terms, capacity is open tills, staffed counters and space devoted to promotions so managers can intervene immediately.

Privacy-first analytics: why LiDAR beats cameras

That push for real-time action is paired with an equally firm stance on privacy. Hyperspace uses LiDAR (Light Detection and Ranging) rather than cameras, and Nestola claims it enables 100 per cent customer privacy by design.

LiDAR emits laser pulses to build a 3D map; it captures shapes and motion, not personal features. In Ulisse’s model, a shopper is a moving cluster of points, a trajectory with no personally identifiable information.

Nestola argues this matters in Southeast Asia as regulators tighten rules on data use, with Singapore’s PDPA a clear signpost. It lets retailers extract sophisticated analytics without compromising anonymity, a requirement for shoppers and for legal teams approving deployments.

With cameras, even if the video is later blurred, the raw footage still exists, creating compliance burdens. With LiDAR, Ulisse says, there is no face, no skin colour, no attribute to record. “Cameras identify individuals,” he says. “LiDAR understands movement.”

Designing for chaos: Southeast Asia’s diverse retail formats

Ulisse also had to design for the region’s diverse range of store formats. Hyperspace is layout-agnostic, built around a 3D Venue Builder and a LiDAR Coverage Planner. Retailers can upload a 2D architectural drawing, typically a DWG file, and the system parses it to generate an accurate 3D digital twin.

If plans do not exist, Ulisse can quickly create the layout using manual tools. Coverage is provided by scalable sensor fusion: multiple low-cost LiDAR units are installed, and their streams are stitched into a seamless view of the venue. The approach is meant to work across extremes, from a 100-square-metre convenience shop to a 10,000-square-metre hypermarket, and from tight city aisles to open-plan big-box floors, without sacrificing tracking quality or operational usefulness.

Operational gains at scale: where small improvements compound

Nestola sees Southeast Asia’s big opportunity as operational efficiency at scale. Retail is intensely competitive, and margins are often thin, so small gains in throughput and revenue per square metre compound quickly.

Hyperspace focuses on two daily drains on profit.

Queue management is first: the platform predicts queue formation and alerts staff to open additional checkouts before lines become long enough to trigger abandonment.

Second is staff and space allocation. By showing where customers are, and where they are not, managers can move staff to the right zones and rework promotions or layouts to monetise underused space. The pitch is practical: improve performance without a major refit so the same store can serve more shoppers each hour.

Also Read: AI shopping adoption surges 39 per cent in APAC, fueling retail tech investments

Ulisse will not name Southeast Asian clients, citing confidentiality agreements, but Nestola says pilots in the region echo work with European retailers such as Italy’s Esselunga. In one deployment with a major grocery chain in Singapore, Hyperspace was trained on checkout operations and the fresh produce section.

Ulisse says predictive alerts helped managers redeploy staff and cut average checkout wait times by 45 per cent during peak hours. The same retailer used foot-traffic and dwell-time analysis to reposition a key fresh fruit promotional display, and Ulisse reports a 22 per cent increase in category sales within the first month.

For him, the takeaway is simple: spatial analytics matters only when it becomes a decision quickly enough to change the customer experience on the floor, not paper.

Turning movement into money: decoding shopper intent

Hyperspace’s intent engine is built on what Nestola calls the “collective physics of shopping”: movement patterns, analysed at scale, become proxies for commercial intent. The system does not try to read individual psychology. Instead, it searches for repeatable signatures across thousands of anonymous trajectories.

A direct, accelerating path to the checkout signals purchase intent. Deceleration and repeated micro-stops in front of a shelf indicate consideration. A rise in dwell time and approach frequency around an endcap display versus baseline suggests promotional pull. When flow speed drops suddenly, or clusters form in odd places, Hyperspace flags friction—congestion, obstructions, confusion—so staff can intervene.

Ulisse says its core LiDAR tracking reaches over 99 per cent accuracy in detecting and continuously tracking shoppers anonymously. At the same time, queue prediction models have shown over 95 per cent accuracy in forecasting wait times—enough, Nestola argues, to act before queues become visible problems.

Measuring what matters: from footfall to causal impact

For in-store media, Ulisse offers PEBLE (Post-Exposure Behavioural Lift Engine), which aims to measure the causal impact of advertising. It compares the post-exposure behaviour of shoppers detected in front of a digital screen with that of a matched control group who were not exposed, an approach Ulisse says has been validated by Deloitte.

Hyperspace is designed to plug into existing retail systems, acting as a central nervous system. DOOH integrations can link an entrance ad for a new drink to later visits to the beverage aisle. POS links correlate traffic and dwell time with sales. Staff-management integrations route alerts to handheld devices, telling teams to open checkouts or assist in specific aisles in real time, store-wide.

Scaling without capex: a service-led business model

Ulisse’s go-to-market is shaped by cash flow realities, particularly for SMEs. Hyperspace is sold as LiDAR-as-a-Service, bundling hardware, software, installation and support into a monthly subscription with no upfront hardware bill. The model preserves capital—zero CAPEX—while delivering a typical payback period of under three months, and it lets retailers scale from one site to many without repeated big purchases.

Also Read: Chaos is a ladder: How instant retail is turning stores into fulfilment powerhouses

Even so, he expects two barriers: perceived complexity and resistance to change. Ulisse’s answer is a “30-minute deployment” playbook, automated floor-plan import and sensor placement, plus plug-and-play Ulisse Box edge servers. The system is framed as augmented intelligence: simple alerts that help managers act faster, not replacements for judgment. It targets post-pandemic shoppers who demand speed and less waiting.

Beyond retail: building a universal operating system for physical spaces

Hyperspace can monitor occupancy and flow to reduce overcrowding and help test new formats, such as dark stores. The system keeps learning as behaviour shifts. On the roadmap is AI Narrator, turning analytics into prompts—flagging that sales are down 15 per cent because an obstruction near the entrance is slowing traffic. Ulisse is focusing on grocery in Singapore, Malaysia and Thailand, then expanding into airports, malls and smart buildings with Kone as a partner. The end goal is a universal operating system for privacy-safe, high-performance spaces.

The post Hyperspace is making stores think and act like websites appeared first on e27.