Machine Learning Engineer

ML/AI Engineer — AlgeriaRemote · Full-Time · KosmetikOn
KosmetikOn is a software company building Labify, an AI-powered platform for the cosmetics industry. We work with large amounts of scientific, formulation, regulatory, and manufacturing data and are increasingly building our own machine-learning and AI systems around this data.
We are looking for a strong ML/AI Engineer based in Algeria to join our engineering team and take ownership of real-world machine-learning and AI systems.
This is a hands-on engineering role. We are looking for someone who genuinely understands machine learning — not someone whose experience is limited to calling LLM APIs or building chatbots.
What you'll work on
You will work across both classical machine learning and modern AI/LLM systems, including:
Machine-learning models for scientific and formulation data
Regression and prediction systems
Non-linear regression and optimization
Neural networks and deep learning
LLM-powered applications
RAG (Retrieval-Augmented Generation) systems
Embeddings, vector databases and semantic search
Fine-tuning and evaluation of language models
AI agents and tool-using systems
Model serving and inference optimization
Data pipelines and ML infrastructure
Production AI systems running on our own infrastructure and data centers
A major part of the role will be taking models from research/prototype → production.
What we're looking forStrong ML fundamentals — mandatory
You should have a deep practical understanding of machine learning, including:
Python
scikit-learn
Linear and non-linear regression
Classification
Feature engineering
Model evaluation and validation
Overfitting / regularization
Optimization
Neural networks
Deep learning
PyTorch or TensorFlow
Understanding of the mathematics behind ML algorithms
You should be comfortable looking at a dataset and deciding which model makes sense, why it makes sense, how to validate it, and how to deploy it.
We are particularly interested in candidates who understand ML beyond simply using pre-built APIs.
Modern AI / LLMs
Strong practical experience with:
LLMs and transformer architectures
RAG architectures
Embeddings
Vector databases
Semantic / hybrid search
Document processing and chunking
Retrieval and reranking
LLM evaluation
Prompt engineering
Fine-tuning / parameter-efficient fine-tuning
Local LLM inference
Open-source models
AI agents and tool calling
Model serving and inference optimization
Experience with technologies such as PyTorch, Hugging Face, vLLM, Ollama, FAISS, Qdrant, pgvector or similar is highly relevant.
Production & infrastructure
You must be comfortable with the fact that our AI systems are not necessarily going to live in the cloud.
We operate our own infrastructure and data centers, and some AI systems need to run completely within our controlled environment.
You should therefore be comfortable with:
Linux
Docker
Git
REST APIs
Production Python services
Model deployment
GPU inference
Monitoring and logging
Performance optimization
CPU/GPU resource management
Networking and security fundamentals
Deploying services on physical or virtual servers
Experience deploying AI systems on-premise is a significant advantage.
You should be able to go from:
"We have a model that works"
to:
"This model is running reliably in production, serving real users, using our own infrastructure."
The ideal candidate
You are probably a good fit if you:
Have a Master's degree in Machine Learning, AI, Computer Science, Mathematics, Data Science, Engineering or a related field
Have several years of hands-on ML/AI experience
Have built and deployed ML systems rather than only studied them
Understand both classical ML and modern LLM-based AI
Enjoy working close to the infrastructure
Can work independently and take ownership of projects
Are comfortable reading papers and turning ideas into working systems
Care about model quality, latency, reliability and maintainability
Are excited by the idea of building AI systems around proprietary scientific and industrial data
Nice to have
Experience with any of the following is a plus:
Scientific or chemical data
Time-series modelling
Optimization
Numerical methods
Computer vision
Recommendation systems
Knowledge graphs
MLOps
Kubernetes
CUDA
Distributed inference
Quantization
Model compression
GPU optimization
LLM fine-tuning
Multi-GPU inference
On-premise AI infrastructure
Location & working arrangement
Location: Algeria
Work arrangement: Fully remote
Employment: Full-time
You will work closely with our engineering team and have significant autonomy over the AI/ML systems you build.
What we offer
Work on real production AI, not experimental demos
Opportunity to build ML systems around proprietary scientific and industrial data
Exposure to both classical machine learning and cutting-edge LLM technology
Direct involvement in architecture, implementation and deployment
Remote work
Long-term position within a growing software company
High level of technical ownership and autonomy
How to apply
Please send:
Your CV
A short description of the most technically challenging ML/AI system you have built
GitHub, portfolio or other relevant work if available
Your expected monthly salary
Important: Please highlight projects where you personally designed, implemented and deployed the ML/AI system. We are much more interested in demonstrable technical ability than a long list of technologies on a CV.

Work arrangement
No