We’re looking for a skilled ML/AI specialist to contribute to the development of an innovative social dining platform based in the UAE. This is a short-term, part-time opportunity (up to one month) with a well-defined scope of work. Your main task will be to design and implement a matching algorithm based on user quiz responses — the goal is to deliver a simple and effective MVP-level solution that runs autonomously on a weekly basis.
What is the project idea?
The platform, SixPlates, connects three couples for shared dining experiences in Dubai and Abu Dhabi. It’s designed to help Expats and Travelers build meaningful social connections through curated restaurant experiences. The platform features a sophisticated matching system and is supported by a dedicated admin team. The product is being built from scratch.
What exciting things is the product doing for the community?
SixPlates transforms the way people connect offline in a new country or city. By intelligently matching couples with similar interests, it creates real-life social connections in premium restaurant settings — fostering community, friendship, and belonging for expats and travelers.
What is the team size and structure?
Our team consists of: Full-stack Engineer, Project Manager, QA, DevOps. You’ll also collaborate directly with the project’s founders on product vision and priorities.
How many stages of the interview are there? * Interview with the Recruiter — up to 30 min; * Technical interview — up to 1 hour.
Requirements: * 4+ years of experience as an AI Engineer, Data Scientist or in a similar role; * Experience with clustering, embeddings, vector similarity; * Experience working with Django REST Framework; * Hands-on knowledge of PyTorch or TensorFlow; * Understanding of transformer models (e.g., sentence-transformers); * Solid skills in PostgreSQL, ElasticSearch, Pandas; * Experience with Docker, CI/CD (GitHub Pipelines); * Familiarity with Azure services: Web Apps, Container Registry; * English — Upper-Intermediate or higher.
Responsibilities: * Design and deliver an interpretable couple-matching algorithm based on quiz data using embeddings, clustering, or similarity scoring techniques; * Build and maintain data pipelines for quiz data and vector embedding storage; * Incorporate user feedback into periodic model retraining; * Collaborate with the backend team to expose matching logic through REST API; * Implement similarity search using Azure Cosmos DB or ChromaDB, aligned with the Azure ecosystem and cost-efficiency goals.
