Key Responsibilities:

Design machine learning models focused on natural language processing and anomaly detection using PyTorch.

Train models leveraging real-time data from Kafka streams and Elasticsearch logs.

Develop AI-generated summaries to be visualised or fed into ARM workflows.

Create FastAPI endpoints to serve ML models for real-time inference.

Continuously test and refine models using real-world SOC data in active feedback loops.

Apply clustering and classification methods to identify cyber threats and behavioural anomalies.

Integrate model outputs with SIEM data pipelines for enhanced threat detection.

Use Docker and MLFlow to manage containerisation and version control for deployed models.

Requirements:

At least 4 years of experience in AI/ML engineering; background in cybersecurity analytics is preferred.

Strong Python development skills with expertise in PyTorch, FastAPI and scikit-learn.

Comprehensive understanding of anomaly detection, clustering and classification techniques.

Hands-on experience with log parsing, SIEM datasets and time-series analysis.

Familiarity with Kafka for stream data processing and Elasticsearch for log analysis.

Knowledge of containerisation and model versioning tools such as Docker and MLFlow.

Ability to build inference APIs for deploying machine learning models in production.

Experience working with feedback loops and model performance tuning using real-world data.