Successful delivery of complex technology projects requires the ability to form deep understanding of both both the semantic structure of the problem domain and the possibilities of the given technology.

Capturing Domain expertise

Machine learning value derives from automating the decision making of experts in a domain.  In order to build effective models, the semantic structure and content of that expertise must be captured in a format suitable for training machine learning models.

This requires simultaneous ability to simultaneously understand complex domain and modelling issues to development appropriate representations.

  • translate unstructured domain expertise to feature labels
  • mapping domain expertise to appropriate modelling solutions
  • product requirements documentation

Ground truth Data labelling

Preparing ground truth data sets is a multi-stage process typically involving the engagement of vendors or crowd-sourcing platforms

  • formalise data requirements for machine learning
  • work with business stakeholders to obtain required data
  • Preparation of semantic categorisation and data labelling
  • Project Management

ML Model product management

ML Product management complements the extremely deep but narrow skillset of ML engineering by managing the requirements, data and outputs required to effectively meet business needs.

  • engage scientists/ developers/engineers technically
  • champion vision for required outcomes


Cloud platform services

Google Cloud Platform

Google Machine Learning APIs

Google AutoML


Keras, Tensorflow


Python, Javascript

Operating Systems

Linux (Ubuntu, Debian), Windows