🌾

Specialise in working with founders to develop proof of concept and minimal viable products that apply machine learning.Working with teams to improve and implement:- Data management.- Gold-standard dataset creation.- Model training.- Deployment and monitoring of models.

Led technical and operational strategy transformation in game development:- Implemented new iteration cycle, streamlining development workflow.- Introduced product & client focused OKR strategic framework within MK.- Enhanced company wide communication, aligning teams with strategic objectives.- Elevated internal build cadence to enable regular fortnightly release testing sessions.- Achieved month-on-month decline or stabilisation in ANR rates across all live titles.- Fostered culture of excellence, targeting 4.0+ star reviews on app stores.- Empowered teams to deliver seamless, engaging gaming experiences.Spearheaded prototyping of live operations systems, including:- Over-the-air content delivery.- Analytics platforms.- Remote configuration for in-development titles.

As Principal Engineer at AIML my role was:- Setting strategic and technical direction for the engineering team at AIML, e.g., shifting the machine learning engineering team towards offering internal services and serving as a profit centre for AIML- Working internally with operations management to align with and help to set the strategic vision for AIML- Working externally with clients to assist with scoping and shaping projects so that they are more likely to be successful (based on data requirements, time, scope, budget, skills and so forth)- Leading our team to deliver several machine learning demonstration systems into State Government - Leadership and technical direction for the machine learning engineering team - setting up coding standards, cluster-based model training and evaluation, data pipeline standard implementations, and other standard practices which are now adopted within the team- Introducing a SCRUM-based agile methodology into the engineering team at AIML- Working closely with internal and external stakeholders with conflicting requirements, timelines, and product maturity requirements – to build and execute on a well estimated roadmap of work- Planning, architecting, and implementing prototype software across many disciplines including computer vision, natural language processing, data analytics and statistical modelling

At GAMURS I was the CTO and Project Lead for the GamerAI project. In this role I was responsible for:- Leadership and technical direction of the engineering and data science team- Working with my software architects to design and implement a scalable solution for the real time analysis of video data from e-sports video games (i.e. Dota2, LoL, Apex Legends, Overwatch, etc.)- Establishing and utilising best-practice techniques for data science workflows, including data management, and building robust model pipelines for training, serving, and model implementation- Working with our product management team to iterate on the system in response to changing customer requirements and user experience research- Working with my architecture team to plan our roadmap and backlog of work in depth to provide estimates of timelines and delivery of features to our marketing team and key stakeholders- Establishing engineering practices and principles within the team, from how we format and test our code, to how our CICD processes automated over the wire updates to our desktop applicationSome of the highlights of the technical achievements that I delivered into the GamerAI products include:- Developing computer vision models using Keras (on Tensorflow) in idiomatic modern Python- Deployment of these models into a .NET Core (F#) application to achieve near real-time analysis of video data- Building a post-match data analysis service in .NET Core (F#) on AppEngine to analyse frame event data (serialised in protobuf) to produce the match and event documents indexed by ElasticSearch that appeared in our user interface- Developed bespoke video capture software, in well-tested modern C++11 code, capable of streaming uncompressed data to our F# desktop application, while remuxing the data into FLV/MP4 videos for upload by our desktop application post-match

Fivecast™ began as Data to Decisions CRC’s Beat the News project with the aim to build an automatic forecasting capacity for law enforcement and national security. A strong collaboration between D2D CRC, law enforcement and national security agencies and several university research groups has led to world-leading forecasting technology – now available as the Fivecast™ product suite.

Product Lead and Software Architect for the "Beat the News" R&D program at the Data to Decisions CRC (D2DCRC). In this role, the team developed a textual-analytics platform that delivered monitoring and discovery of population-level events; such as social disruption, political crises, and disease outbreak.Working closely with partners - the team were able to take this application from Proof of Concept to user-agency trials.I frequently engaged with our end user agencies to identify the key qualitative insights that we can compute and display through our platform. I also worked with my engineering team to design and build software following best-practice Agile development methodologies and deliver our capabilities using a 'cloud-first' approach.

Helped to build a capability for continuous, automated analysis of publicly available data in order to anticipate and detect significant societal events such as such political election outcomes or disease outbreak.Worked closely with University partners to understand the research issues and help guide prototype implementations. Provided leadership within our engineering team to implement solutions that facilitate a collaborative environment and scale-out implementations the research prototypes.Built our Data Science Platform using technologies such as: Apache Spark, HBase, HDFS, YARN, Python, Java and Scala.

Technical lead on large scale JEE project. This involved re-factoring the existing Maven project POMs to improve project structure and build time; and removing duplicated or unused dependencies. I worked with international partners to improve the application and bring it into a configuration managed, easily deployed and maintainable state. I also coordinated with contractors and partners to capture knowledge about the legacy systems we were improving on and fed this knowledge back to the Adelaide development group.I was the Chair of the Continuous Integration and Deployment Special Interest Group at iiNet. With this role I was working towards developing a standard for our Development Infrastructure to facilitate CICD.Actively promoting and taking part in Scrum-based Agile Methodologies to deliver incremental improvements to systems within iiNet.

Research and Development of software tools for computer security. This includes using Hadoop for data analytics, applying machine learning and presenting results using information visualisation.Undertook many training courses in machine learning, windows kernel driver development and computer security.Implemented memory forensics techniques for virtual infrastructure and implemented visualisation dashboard for computer security data.

Helped to develop software for the analysis and presentation of metadata extracted from common forensic toolkits including EnCase, FTK, XRY and Cellebrite. This used Java and OpenGL to create information visualisations. The tool was backed by a MySQL database. We worked with AFP and PMC to gather requirements for the tool.Developed Wireless Forensics Analysis Toolkit (WiFAT). This was developed in C using libpcap to passively scan wireless signals and extract digital evidence (metadata only to comply with the Australian Telecommunications Act). Developed a Java-based server to collect the data into a sqlite database. The server enabled many drones to send their data back for analysis.Helped to create and organise a digital forensics course for students from Mississippi State University. Also conducted research and reporting on the viability of using VMware platforms within Australian Customs. The goal was to determine if VMware systems could be recognised from external sites (such as forums).

Development of Software Tools to analyse the use of various techniques in the visualisation of radar data.Worked towards the development of a new Software Development Environment and testing solutions as part of JORN.