Senior AI Engineer - Multilingual LLMs for Africa

Awarri; United States; Remote; USD 120,000 - 180,000; Full time

Job summary

Awarri is seeking an AI/ML engineer to lead the design, training, and optimization of large-scale language models for African languages. You will work on data preprocessing, tokenization, and embedding for low-resource datasets, and fine-tune transformers for text generation, translation, and sentiment analysis.

You will collaborate with linguists, deploy in cloud environments, and build robust evaluation frameworks using metrics such as perplexity and BLEU.

Qualifications

• Minimum of 3 years in AI/ML engineering focusing on NLP/LLMs.
• Strong expertise with transformer models (GPT, BERT, T5).
• Proficiency in TensorFlow, PyTorch, or Hugging Face Transformers.
• Experience training and fine-tuning LLMs on large multilingual data.
• Strong programming skills and NLP library expertise.
• Experience building text data collection & annotation pipelines.
• Familiarity with AWS, Azure or GCP.

Responsibilities

• Lead design, training, and optimization of language models for African languages and dialects.
• Develop data preprocessing, tokenization, and embedding for low-resource datasets.
• Fine-tune transformer architectures for text generation, translation, and sentiment analysis.
• Design scalable pipelines for data collection, annotation, and preprocessing.
• Collaborate with linguists to ensure alignment with cultural sensitivities.
• Deploy, monitor, and maintain production LLM systems in cloud environments.
• Establish evaluation frameworks using perplexity and BLEU metrics.

Skills

NLP & LLMs

Transformer models

TensorFlow / PyTorch

Hugging Face Transformers

Multilingual NLP

Cloud computing

Tools

Hugging Face

Job description

Awarri is seeking an AI/ML engineer to lead the design, training, and optimization of large-scale language models for African languages. You will work on data preprocessing, tokenization, and embedding for low-resource datasets, and fine-tune transformers for text generation, translation, and sentiment analysis.

You will collaborate with linguists, deploy in cloud environments, and build robust evaluation frameworks using metrics such as perplexity and BLEU.