Verified Peer Reviewer

Finetuning Large Language Models by Lamini

Pretraining LLMs by Upstage

Preprocessing Unstructured Data for LLM Applications - Unstructured

LLMops - Google Cloud Platform

Multi AI Agent Systems with crewAI - crewAI

Evaluating and Debugging Generative AI - Weights & Biases

Building Systems with the ChatGPT API - OpenAI

Serverless LLM Apps Amazon Bedrock

ChatGPT Prompt Engineering for Developers - OpenAI

Master of Business Administration - MBA

Psychology (Honours)

Doctor of Philosophy (Ph.D.)

Doctor of Philosophy (PhD)

Developed cloud architecture and data analysis pipelines for high-resolution satellite imagery analysis at Source Localisation. Led the team in implementing innovative solutions to detect mineralisation deposits efficiently. Achieved significant cost savings and increased productivity through streamlined processes and automation.

Chief Technology Officer

Developed pipelines for the automatic analysis and visualisation of supply chain logistics, inventory management, procurement, sales, and delivery. Through detailed product and market analytics was able to successfully increase conversion rate, retain lifetime customers, and reduce churn. Implemented ML and AI-driven advertising campaigns through Google, Facebook, Influencer Marketing, and SEO.

Head of Data Science

Designed research studies to combine lab group's behavioural paradigms with imaging techniques for real-time brain imaging. Compiled experimental and psychometric data, archived functional neuroimaging data, and co-founded Sydney Neuromethods Group. Set up fNIRS laboratory from scratch, coached associates on behaviour scoring, and drafted reports for students. Published papers on the effect of using computer simulations to help students more effectively learn principles of complexity theory. Member of School of Psychology, Education, Brain and Mind Centre, and Learning, Cognition and Brain Sciences Unit.

Postdoctoral Research Fellow

Compiled and organised data routines for functional neuroimaging analysis in sleep research. Analysed continuous-wave functional near infrared spectroscopy data through custom code. Developed software to operate custom hardware and analyse sensor-data through ML algorithms to extract meaningful insights regarding brain function from complex and noisy time-series datasets collected while patients were asleep.

Senior Data Scientist

Visiting scientist of Professor John Duncan. Performed an experiment analysing the effect that organising objects in spatial locations can facilitate performance at complex goal-directed tasks through providing a cognitive scaffold that forced participants into "chunking" components of the task. 

Research Scientist

I designed studies to investigate the neurocognitive development occurring over middle childhood with the use of neuroimaging techniques such as EEG and fMRI. I founded the West Australian Neuromethods Group to support researchers using neuroimaging methodologies and supervised a large number of postgraduate students conducting research on brain electrophysiology.

Boutique data consultancy firm that specialises in statistical consulting and the analysis of complex datasets. Projects range from the implementation of large deep learning AI vision and language models on the cloud and edge IoT devices to experimental design and analysis of pharmaceutical drug trials.

Senior Data Specialist

Led the development of an AI-assisted insurance underwriter using large language models for enhanced pre-assessment handling. Implemented innovative strategies to increase efficiency and specificity in insurance processes. Achieved significant improvements in processing time, resulting in a more streamlined insurance workflow.

Head of Artificial Intelligence

<p>Acting as the Head of AI Engineering at Comtrac, I led the development of incorporating AI solutions into a software product for conducting investigative research. By implementing generative AI and natural language processing techniques, key pieces of client's evidence were able to be extracted, summarised and transformed. This improved efficiency and accuracy in investigative processes, resulting in significant time and cost savings for Comtrac's clients.</p>

Head of AI Engineering

<p>As an AI Consultant, my leadership extends to developing containerized offline Large Language Models (LLMs) utilising clients' internal company data. Prioritising performance, privacy, and security, I facilitate the implementation of sophisticated RAG pipelines, LLM-router mechanisms, and multi-agent strategies to boost model efficacy. My work also involves quantization of large LLMs and the automated fine-tuning of prompts for smaller LLMs tailored to specific business cases, resulting in highly performant and accurate bespoke models.</p>

MLOps

AI Implementation

Kubernetes

Deep Learning

Python

MongoDB

Django

JavaScript

Machine Learning

AI Consultant

Head of AI | PhD (Cognitive Neuroscience / Artificial Intelligence) | Cognitive Computing | AI Models & Infrastructure

projects

substack

medium

github

devto

productHunt

Know about Gareth Roberts, PhD’s work experience, education, bootcamps, certifications, and much more. Gareth Roberts, PhD