Senior NLP Engineer (AI)

Location: On-site at the employer's office in Almaty, Kazakhstan, 22 Kazhymukan Street.

Company

Verigram is an international AI company developing digital identity and fraud prevention solutions for banks, fintechs, and online platforms.

What you will do

• Design, train, fine-tune, and evaluate machine learning and deep learning models for conversational AI, NLP, computer vision, and related AI architectures.
• Build datasets through cleaning, labeling, augmentation, and synthetic data generation; define experiments and evaluation metrics.
• Improve model accuracy, robustness, latency, and computational efficiency.
• Research new models, architectures, papers, and AI technologies, translating relevant ideas into practical implementations.
• Work with ML Engineers to productionize models, build reliable inference pipelines, and support monitoring, retraining, and continuous improvement.
• Partner with business stakeholders to identify AI/ML opportunities, define success metrics, and contribute to architecture decisions, code reviews, documentation, and ML best practices.

Requirements

• 5+ years of experience in Machine Learning, Data Science, Applied AI, or related fields.
• Strong knowledge of machine learning, deep learning, probability, statistics, and optimization.
• Strong Python skills and practical experience with PyTorch or a comparable deep learning framework.
• Experience training and fine-tuning neural networks and working with modern architectures such as Transformers, LLMs, and VLMs.
• Experience reading research papers and implementing relevant ideas in practice.
• Experience shipping ML models to production and understanding the full ML lifecycle.
• Practical experience building AI agents and LLM pipelines with LangChain or LangGraph.
• Experience with Git, Linux, Docker, and common ML development workflows.
• Strong communication skills, including explaining technical decisions, translating undefined business objectives into technical plans, articulating architectural decisions, and working independently.

Nice to have

• Experience building high-load services, backend systems, or GPU inference solutions.
• Experience with LLMOps principles, including CI/CD, orchestration, and prompt version control.
• Understanding of LLM application security.
• Experience developing and launching BERT, Sentence-BERT, T5, fastText, word2vec, or similar architectures.
• Experience with experiment tracking tools such as MLflow, ClearML, Weights & Biases, or similar.

Culture & Benefits

• Work on real production ML and LLM projects used by thousands of users.
• Access to GPU infrastructure.
• Strong influence on AI product architecture.
• Strong engineering culture with low bureaucracy.
• Competitive salary.