Company Overview

Yakeey is a fast-growing PropTech company headquartered in Casablanca, Morocco. Our mission is to simplify and accelerate real-estate transactions by connecting buyers, sellers, renters and advisors through a single digital platform.
Yakeey operates a multi-product ecosystem including a property marketplace, a credit application platform, and YakeeyVal, our in-house property valuation engine, built on data, algorithms, and AI/ML.

Role Summary

As a Graph-Based Modeling Intern, you will explore graph neural networks (GNNs) as a potential alternative to heuristic spatial smoothing techniques currently used in YakeeyVal. This role is research-oriented, with a strong emphasis on experimentation, benchmarking, and critical evaluation.

Key Responsibilities

• Model spatial entities (city blocks, neighborhoods) as graphs

• Define node and edge features using transaction and spatial data

• Implement a GNN prototype for property price estimation or smoothing

• Compare GNN results with existing spatial smoothing logic

• Analyze robustness, generalization, and interpretability

• Provide clear recommendations on feasibility and next steps

• Strong Python skills

• Basic understanding of graph theory

• Familiarity with PyTorch or similar frameworks

• Interest in Graph Neural Networks or geometric deep learning

• Comfortable reading academic or technical papers

• Bonus: prior exposure to GNN libraries (PyG, DGL)

• First conversation with the HR team

• Practical case study (graph modeling or reasoning task)

• Technical interview with the AI/ML team

• Final meeting with management (if needed)