Day-to-Day Responsibilities
Collect, clean and label text or speech data, and prepare it for model training.
Pre-process text using tokenisation, stop-word removal, lemmatisation and text normalisation.
Build and fine-tune models for tasks such as text classification, sentiment analysis and named entity recognition.
Train and evaluate transformer-based models such as BERT and GPT-style models, using metrics that match the product goal.
Develop chatbots, question answering systems and summarisation tools, and connect them to products through APIs.
Work with product teams, linguists and data annotators to understand user needs and fix problems such as wrong intent detection or biased answers.
Monitor models in production, track performance drift and retrain them when the way people use language changes.
