Optimize the visibility of real-time AI data processing

Bringing transparency and spatial visibility to the visual data automation process.

Skills

UX/UI Design, SaaS B2B

Sector

Industry, Vision AI

Date

2023/2024

Mockup of the user interface for the data integration module.

Context

Data Flow is the visual data integration module of the Alteia platform. It automates the process of transforming raw files – images, 3D models – into structured and usable information.

The problem

Opacity of the process

Lack of clear information on the status of data ingestion and processing. Users did not know the progress of their pipeline.

Lack of spatial visualization

Users struggled to understand the precise geographic location of ingested visual data due to the lack of a map interface.

Objectives

Develop a comprehensive monitoring dashboard providing real-time visibility into ingestion processes.
Integrate an interactive 2D mapping interface to visualize and manipulate geospatial data.

Ideation process

Stakeholder Interviews

Close collaboration with the COO, an expert user of the module, to gather firsthand insights and identify critical pain points.

"How Might We" Workshop

Session with the Product Owner to reframe problems as opportunities by brainstorming creative and actionable solutions.

MoSCoW Methodology

Categorizing features into Must-have, Should-have, Could-have, and Won't-have to align efforts with the most critical needs.

Key screens

Overview Tab

A summary view with key metrics (completed, in progress, and error-prone files) in a donut chart format. The second section details each status with direct links to the associated files.

Files Tab

Status tracking for each ingested file (import, contextualization, quality check, export). For geospatial data, the footprint of each file can be viewed directly on the map.

File table with columns for Name, Status, Contextualization status, Quality check status, and Import date, accompanied by an interactive 2D map displaying the cyan rectangular geospatial footprint of each file.

Assets Tab

Management of final data processing results: list of resources with coverage, percentage, and version. Visualization of each asset's footprint on the map to understand its geographic scope.

Assets interface with a table listing Type, Coverage, Coverage %, Version, and a 2D map showing a green and red power transmission line following a linear corridor with a detailed geospatial footprint.

Results

Feedback from stakeholders and users has been overwhelmingly positive, confirming the usefulness and impact of the changes implemented.

Through internal interviews, we were able to demonstrate to C-level executives the crucial importance of involving users in the design process, even when this practice is not yet well established within the company.

Lessons learned

User research is always possible

It is entirely possible to conduct user research with available resources, even when this practice is not yet ingrained in the company culture. Every insight has value.

Involve users from the start

Collaborating directly with the COO - as an expert user - allowed for the rapid structuring of priorities and avoided costly back-and-forth at the end of the project.

Defend the value of design

Presenting the research results to senior management helped solidify the legitimacy of the UX approach within the organization, paving the way for a more mature design culture.

To embed a website or widget, add it to the properties panel.