We are looking for an experienced Applied Scientist with strong foundations in machine learning, deep learning and model development. The ideal candidate should have hands-on experience building, training, evaluating and improving machine learning models, with the ability to reason about model architecture, data requirements, algorithm selection and performance trade-offs.

This role is suited to someone who understands the science behind machine learning models and can make informed technical decisions throughout the model development lifecycle. The candidate should be comfortable working with data, experimenting with different approaches, evaluating model performance and solving complex problems where the most suitable approach may not be known upfront.

A strong understanding of neural networks, transformer architectures, embeddings and model training techniques is important. Experience limited to integrating existing AI APIs, using pre-trained models without deeper understanding, or building applications around LLMs alone will not be sufficient for this role.

Key Responsibilities

• Design, develop and improve machine learning models to address business and technical problems.
• Analyze datasets, identify relevant features, handle data quality issues and prepare training, validation and test datasets.
• Select appropriate algorithms and model architectures based on the problem, data characteristics and expected outcomes.
• Train, tune and optimize machine learning and deep learning models, with a clear understanding of the decisions behind the chosen approach.
• Apply core machine learning concepts, including supervised and unsupervised learning, feature engineering, regularization, bias-variance trade-offs and model generalization.
• Work with neural networks and deep learning techniques, including an understanding of loss functions, optimization methods, training stability and overfitting.
• Demonstrate a strong understanding of transformer architectures, attention mechanisms, embeddings and the considerations involved in training or fine-tuning models.
• Establish appropriate evaluation methodologies, select relevant metrics, design experiments and investigate model errors and failure cases.
• Compare alternative modelling approaches, interpret experimental results and recommend improvements based on evidence.
• Conduct structured experimentation to understand model limitations, performance trade-offs and opportunities for improvement.
• Collaborate with engineering and product teams to translate business problems into measurable machine learning solutions and support their practical implementation.

Required Skills and Experience

• 5–8 years of relevant experience in machine learning, applied science, data science or AI model development.
• Strong programming skills in Python and hands-on experience with relevant ML libraries and frameworks.
• Strong understanding of fundamental machine learning algorithms, model selection, feature engineering, training techniques and evaluation metrics.
• Demonstrable experience building and training models, including decisions related to data preparation, algorithm selection, model tuning and performance improvement.
• Practical knowledge of deep learning concepts, neural network architectures, loss functions and optimization techniques.
• A solid understanding of transformer architecture, attention mechanisms, tokenization and embeddings.
• Experience evaluating models using appropriate metrics, experimental methods, error analysis and validation techniques.
• Ability to investigate model performance issues, identify root causes and propose technically sound improvements.
• Strong analytical and problem-solving skills, with the ability to compare approaches and explain the reasoning behind technical decisions.
• Ability to clearly communicate modelling choices, experimental findings, limitations and trade-offs.

Good to Have

• Experience working with large-scale datasets or computationally intensive model-training workloads.
• Hands-on experience training or fine-tuning transformer-based models for specific use cases.
• Familiarity with representation learning, embedding generation, similarity-based methods or information retrieval.
• Experience with experiment tracking, reproducible model development and ML workflow tools.
• Experience taking models from experimentation towards production deployment in collaboration with engineering teams.
• Research publications, technical projects or other demonstrable work in applied machine learning.

What We Are Looking For

We are looking for a candidate who can explain not only which model or framework they used, but also why they chose it, how they trained it, how they evaluated it and what they changed when the results were inadequate.

Candidates should be able to walk through at least one substantial model-development project, covering the original problem, dataset, modelling approach, training process, evaluation methodology, challenges encountered and measurable outcomes.

Candidates whose experience is primarily limited to prompt engineering, API integration, orchestration of existing LLMs or application development using pre-trained models—without strong underlying ML and model-development knowledge—may not be suitable for this track.