Logo du laboratoire
LaCDIA
AI Services & Solutions

Applied Research and Technology Transfer in Artificial Intelligence

LaCDIA develops AI methods and prototypes, then transfers them to real-world use cases in collaboration with its partners.

Our solutions are built on research areas, interdisciplinary projects, and rigorous scientific validation.

Scientific Architecture 2026–2030

Two research axes, one valorization hub

“The collaborative development of scientific and technical skills in artificial intelligence and data science for the sustainable development of the Caribbean.”

AXIS 01

Foundations, Methods and AI Robustness

The methodological backbone of the Laboratory — designing methods that allow systems to learn, reason and generalize robustly, including from scarce, noisy or heterogeneous data.

Self-supervised learningRobustness & generalizationFrugal modelsExplainability & auditBias reductionReasoning under uncertainty

Priority objectives

Develop efficient algorithms with limited data· Study out-of-distribution robustness and generalization· Design models suited to limited computational resources· Develop explainability methods and measure biases

AXIS 02

AI for Complex Data and Real-World Systems

How to adapt, combine and evaluate AI methods when confronted with real-world complexity — multimodal data, dynamic systems, variable-quality environments.

Multimodal data & fusionDetection & segmentationDecision supportModeling & predictionField validationCross-domain transfer

Priority objectives

Learn from scarce, noisy, incomplete data· Develop contextualized decision support systems· Evaluate systems in real conditions· Integrate domain knowledge

Feedback loop

Real-world cases identified by the Hub feed back as new scientific problems for Axes 1 and 2.

HUB

Scientific Valorization and Training

The interface between the Laboratory and the socio-economic world. The Hub valorizes and transfers knowledge from Axes 1 and 2 to organizations, while feeding back real needs as new scientific problems.

Application areas

HealthAgricultureEducationEnvironmentPublic sectorExpertise & prototyping

Application sectors (health, agriculture, education, environment…) belong to the Hub — not to axis titles — preserving a general, lasting scientific architecture.

Transversal requirement

Ethics · Explainability · Security · AI Governance

AI ethics is not a separate axis: it permeates both axes and the Hub, according to the nature of the subjects treated. Each project integrates bias assessment, usage limit documentation and, where applicable, an explainability protocol adapted to its end users.

Bias assessment & mitigation
User-adapted explainability
Data security & cybersecurity
Governance & method traceability
Ethical & regulatory validation
Usage limit documentation

Our approach

A scientific, rigorous, and impact-oriented approach.

1

Diagnosis

Assessment of data, business needs, and field constraints.

2

Modeling

AI design, rapid prototyping, and scientific validation.

3

Deployment

Integration, support, and impact measurement.

AI Solutions

Concrete solutions to analyze, automate, decide, and make information accessible.

chart

Analysis & Prediction

Analyzes past data to detect trends, forecast future situations, and identify anomalies.

chat

Intelligent chatbots

Lets users ask questions in natural language and get immediate answers.

document

Intelligent document access

Uses artificial intelligence to query internal or private documents as if talking with an expert.

network

Intelligent assistant coordination

Coordinates several specialized intelligent assistants to solve complex problems.

automation

Intelligent automation

Automates repetitive or time-consuming tasks while keeping human oversight for critical actions.

settings

Custom AI

Adapts artificial intelligence to your documents, vocabulary, and business context.

Use Cases

Intelligent assistant coordination for agriculture

Supporting agricultural decision-making in real-world conditions.

A team of assistants analyzing field data and observations to propose recommendations.

  • Better agricultural decisions
  • Reduced losses
  • Adaptation to local context

Applied AI flow

Field data -> Knowledge -> AI models -> Decision

1
Collection
2
Analysis
3
Recommendation
4
Field monitoring

Sectors of Application

Solutions adaptable to any sector with data and documents.

Banks and financial institutions
Commerce and retail
Agriculture
NGOs and international organizations
Health
Public sector
Industry
Education

Collaborate with the laboratory

Institutional partnerships, internships, funding, or applied projects: let's build impactful solutions together.

Propose a partnershipApply for an internship