Digeiz Web App : Core Application of Digeiz Back-Office

Digeiz develops an AI that leverages computer vision to provide shopping centers with real-time insights on digital screen audiences and visitor flows, measuring demographics (age and gender) and engagement (store visits, dwell time, screen-to-store conversion…).

Their analytics empower clients to refine marketing strategies and enhance customer experiences by optimizing retail space and media effectiveness.

I have worked on the complete Digeiz back-office value chain with a focus on the Digeiz Web App, which relies on 2 branches:

• Map Editor, the core and most complex application. It is used to create venue maps for deploying Digeiz’s footfall metrics AI.
• The customer dashboard, used to easy to read and to use data to the customers.

Venue maps are at the core of Digeiz’s value chain and serve multiple purposes:

• Providing customers with accurate estimates and defining the project perimeter.
• Coordinating the installation and maintenance of the site’s CCTV camera network.
• Managing site parameters (e.g., contracts per store, escalator directions, digital screen positions, sizes, and fields of view).
• Positioning and maintaining all AI visitor counting stations.
• Creating heatmaps for customer data visualization in the dashboard.

---

My Role & Responsibilities

• Conducted UX research across company stakeholders.
• Led the UI design of the application (visuals withheld for confidentiality reasons)
• Managed the full web app roadmap and wrote specifications.
• Oversaw the development team and ensured smooth execution.
• Trained and managed a junior product manager
• Built a company and product knowledge database.

---

Map Editor is not just a visualization tool—it is the foundation of Digeiz AI’s accuracy.
Since AI-driven footfall analytics depend on real-world spatial mapping, the system must process complex site parameters like:

• Camera positioning & visibility range
• Store layouts & entry points
• Multi-floor escalator traffic flow
• DOOH screen exposure areas
• …

However, site deployment, operation, and maintenance also required periodic updates to account for changes over time, including:

• Camera status (relocations, deactivations, adjustments)
• Counting station updates (new installations, recalibrations)
• Store activity & brand changes (tenant shifts, closures, renovations)
• …

To maintain AI accuracy, these evolving parameters had to be manually updated or periodically checked/refreshed within the system and stored in a structured history.
A key challenge was designing a scalable data model that could integrate these historical updates into AI processing workflows—without requiring real-time tracking—while ensuring consistency and usability over time.

This required translating real-world changes into structured metadata that could be leveraged effectively for both site deployment and AI-driven analytics.

---

Key Challenges

• Creating a new product (almost) from scratch.
• Structuring the Front-End development team and strengthening cross-team collaboration.
• Translating complex real-world parameters into structured data and metadata usable by AI and the system (spatial & temporal considerations).
• Designing a solution adapted to all stakeholders, featuring a scalable ACL architecture to manage different user roles and permissions.
• Delivering fast and continuous updates to the Operations Team, which relied heavily on the tool for site studies and deployments.
• Continually training teams and documenting the product to keep pace with the rapidly evolving web app and ensure it could be used to its full capabilities.

Impact & Results

• Developed an application essential and foundational to Digeiz’s operations and growth.
• Reduced study and deployment time from several weeks to just a few days.
  • Achieved an estimated gain of ~€20k per mall study (-60% time reduction).
• Structured the company lexicon and knowledge database, solving critical alignment issues between teams.
• Implemented versioning and document export systems, used by on-site ops and as key annexes for contracts.
• Delivered a strong « WOW! » effect for customers and prospects.
• Enabled the study, deployment and management of 38 shopping centers (15 active at the time of my departure).
  • Each shopping center represented an average of €100k ARR.