Poitiers · Montreal
We design Data & AI solutions tailored to your business, integrate them into your information system and run them over time.
Your data stays under your jurisdiction, in Europe or in Canada, depending on your regulatory and hosting constraints.
DPLIANCE supports companies and public organisations at every stage of their Data & AI projects: identifying the right use cases, building the solution, putting it into production and maintaining it over time.
Identify the projects that genuinely deserve to be built.
We analyse your processes, your data and your constraints to identify the use cases where Data or AI can deliver a measurable gain.
By the end of the scoping phase, you know what is worth building, with which data, under which regulatory constraints and in which order of priority.
Build a solution that fits your business, not the other way round.
We design and develop applications that integrate with your existing tools and processes.
Document extraction and analysis, document assistants, workflow automation, classification, prediction, segmentation or line-of-business applications: the technology is chosen to fit the problem to be solved.
Explore our expertiseKeep the solution working over time.
A Data or AI solution does not end at go-live. Data changes, usage shifts and models need to be monitored.
We handle monitoring, maintenance and continuous improvement of the solutions we build. When something goes wrong, you talk directly to the engineers who know your project.
Talk to an engineerWe build our own software and support public and private organisations on their Data & AI projects.
For the Dis-ADEME challenge, we built a solution able to analyse funding applications, check their completeness and support their pre-screening. After winning the challenge, DPLIANCE was selected to industrialise the solution.
Read the ADEME case study
Understand your audience without building individual profiles 
Mirage Analytics is our privacy-respecting audience measurement solution. It analyses traffic and browsing behaviour without cookies or fingerprinting, and without depending on an advertising ecosystem.
Discover Mirage Analytics
Audit a website's compliance in a few minutes 
Complio analyses the main visible elements of a site related to its compliance: cookies, trackers, forms, legal notices, privacy policies and data transfers. The tool reports the issues it finds in a structured, prioritised report so DPOs and their clients can act on them.
Discover Complio
Manage consent simply and verifiably Cookilio only loads non-essential services once the visitor has agreed, and issues a receipt for every choice. We built it so that declining really means declining, with a banner that is simple to install and to administer.
A Data or AI project should not start with several months of development.
We move forward step by step, on your data and under your real conditions, so that each stage is validated before investing in the next.
30 minutes
We look at your need, the data available, the technical constraints and the applicable regulatory framework. The goal is to quickly identify what is worth exploring further and the main feasibility conditions.
4 to 8 weeks
We build a first version on a representative sample of your data. You can then assess the results on your own use case before committing to industrialisation.
depends on the project
The solution is integrated into your information system, secured, documented and deployed in the environment that fits your constraints. Hosting can be in Europe, in Canada or on your own infrastructure.
over time
We monitor the solution, fix issues and evolve it alongside your usage. The engineers who built your solution stay involved after go-live.
No data leaves Europe.
A prototype is only worth something if it can become a tool people actually use. From the outset we think about integration, security, monitoring, running costs and maintenance.
The architecture is defined around your constraints: European hosting under GDPR, Canadian hosting when the project falls under Law 25, or deployment on your own infrastructure. The choice of hosting and technology is part of designing the project, not a correction bolted on at the end.
We favour open, reversible architectures. For work built specifically for your organisation, ownership and reversibility terms are agreed at the start of the project, to avoid any unnecessary dependency on DPLIANCE or on a proprietary platform.
A single assessment used to take nearly 12 hours of data handling and writing. DPLIANCE built us a custom tool: we gained 30 to 50% of the time spent on reports, which adds up to 43 days a year. Today I put a real platform in my clients' hands. It signals quality, and it improves our commercial results too.
Nicolas Rolle-Milaguet
Managing Director — STRATÉNERGIE
We were using Google Analytics, which had become too complex for what we needed. With Mirage I collect my main indicators in 30 minutes a month: sessions, read rates, engagement — it is all clear and in one place. For a small communications team that has to report without spending hours on it, this is exactly what we needed.
Amélie N.
Head of Communications — Neoloji — Technopole Grand Poitiers
We mainly work on projects where data can automate, assist or improve a business process: document analysis, internal assistants, workflow processing, prediction, classification, data exploitation or line-of-business applications.
We can step in at the thinking stage or take over a project already under way.
It depends on the project.
Our own products are hosted in Europe. For custom development, we define the architecture around your jurisdiction and your constraints: Europe, Canada or internal infrastructure.
That choice is settled at design time and documented with you.
The budget depends mainly on the data available, the complexity of the problem and how deeply the solution integrates with your information system.
We work in stages so that feasibility and results are validated before the larger development work is committed.
A first 30-minute conversation with an engineer is usually enough to determine the scope worth studying.
For work built specifically for your organisation, ownership of the code, the components involved and the reversibility terms are set out contractually at the start of the project.
Our aim is that you keep control of your solution and of how it evolves.
A sovereign AI agency designs and ships solutions while keeping control of where the data sits, which technical suppliers are involved, which models are used and what the reversibility terms are.
At DPLIANCE, the architecture is defined according to the framework that applies to the project: GDPR in Europe, Law 25 in Quebec, or your organisation's own constraints.
Learn more about our sovereign AI expertiseCompliance does not depend on the supplier alone, but on the architecture of the whole project.
Before starting, you need to know where the data will be hosted, which subprocessors can access it, which models will be used and how the solution could be taken over or migrated.
DPLIANCE has been working on these questions since 2019 and also builds software used by more than 500 organisations in the compliance field.
The cost depends on the use case, the data available, the depth of integration and the security or compliance requirements.
We usually break projects into several stages so their value is confirmed before industrialisation is committed.
A prototype on representative data generally takes several weeks; the go-live schedule then depends heavily on the information system to integrate with.
Yes. Mirage is designed for the CNIL consent exemption, with no special configuration.
It uses no third-party cookies, no fingerprinting and no persistent identifiers, and IP addresses are anonymised at collection.
You can therefore measure your audience without showing an analytics banner.
Discover Mirage AnalyticsMirage answers a different need.
It brings together in one tool what others sell as four: analytics, session replay, conversion optimisation and SEO. It all fits on a dashboard you can understand without training.
The data is hosted in France, at Scaleway, and is not used to feed an advertising ecosystem.
Compare the two approachesOur products are hosted in Europe.
For custom projects, hosting is defined according to the client's constraints: a European environment, a Canadian one or internal infrastructure.
The choice of hosting, suppliers and models is part of the technical scoping of the project.
A process to automate? Data that is hard to exploit? An AI project to scope? Present it directly to an engineer. The first conversation lasts 30 minutes and commits you to nothing: we go through the need, the data available and the main technical or regulatory constraints.
A reply within 24 business hours, from a real person. No commitment.
