Expertise

Supervised, Semi-Supervised, and Unsupervised ML approaches

• Shallow models (Feed-Forward single-layer Neural Networks)

• Deep models (Convolutional and Recurrent)

• Rule-based models (Decision Trees and Association Rules) 

• Ensemble models (Random Forest and Boosting)

• Multi-Task and Transfer models (Kernel-based Shared and Independent Representations)

Planning and Scheduling

• For a simple problem, Mixed Integer Programming with associated optimiser 

• For more complex problem, Answer Set Programming (ASP, WASP) and Planning Domain Definition Language (PDDL, PDDL +) with associated solvers 

Agentic AI

Explainable AI

• For simple problems, natively explainable models (e.g., Decision Trees and Linear models)

• For more complex problems, less-interpretable models (e.g., Deep and Kernelled)

AI Fairness

• Assessment of fairness of AI-based applications

Large Language Models (LLMs)

• LLMs for code generation

• LLMs for asset maintenance and management (intelligent assistant)

Multimodal LLMs

Ingestion

• Extraction, Transformation and Loading (ETL) pipelines to ensure data quality while taking care of security and privacy employing frameworks like Apache Kafka, Apache Flume and Tableau Prep.

Storage

• Data storage solutions to ensure data security and safety, using enterprise level relational and NoSQL database engines (e.g., Oracle Database, PostgreSQL, MySQL, Cassandra, Hbase, etc.).

Analytics

• Data analysis tools to extract useful information from data to support decision making utilising state of the art frameworks and libraries like Pandas, scikit-learn, R, Spark-ML, etc.

Graphics / Dashboards

• Graphical tools to develop dashboards (e.g., Grafana, DASH)

Foundation models

• LLMs (e.g., Llama, Gemini, ChatGPT, etc.)

Dynamic Risk Assessment (DRA) for safety and security (including cyber)

• Management of cascading effects in complex systems

• Intangible costs

• Treatment of High Impact Low Probability (HILP) events