Oxford Physics Graduate, Physics PhD, Imperial MBA Graduate, 770 GMAT (99th Percentile)

Awarded Imperial Innovations Graduate Assistantship in Physical Science/Technology Transfer. Scored 770/800 on the GMAT - in top 1% of candidates globally for business school entrance test, 2003.

Gravimetric, Neutron Scattering and Computational Studies of the Sorption and Diffusion of Gases in Zeolites

Upper Second class BA (Hons) degree in Physics, 1989. Promoted MA by seniority, 1997. Astronomy Honour Moderations paper, 1987; third year options, Astrophysics, Atmospheric and Oceanic Physics

As Chief Scientist of Prometheus Ultraintelligence, I am leading a research program to create a safe and robust path to I.J. Good's concept of "ultraintelligence" as described in his seminal 1965 paper "Speculations Concerning the First Ultraintelligent Machine" . Our work moves beyond simply scaling current AI paradigms. We directly address the core limitations of modern systems: the reliance on correlation instead of true causal reasoning and the brittleness of external, behavioral alignment techniques. Project Prometheus differs from standard foundation models by pursuing a "renewed programme" that synthesizes Good's foundational principles with modern engineering. Instead of correlational attention and external RLHF, we are building an architecture with intrinsic causality and internal governance. Our goal is a system capable of stable, continual learning and recursive self-improvement, grounded in a verifiable, architectural approach to the alignment problem. My activities include: * guiding the program's core thesis: shifting from learning statistical correlations to building systems that operate on causal models of the world. * overseeing the development of our novel "Causal Agentic Mesh (CAM)" architecture, which uses dynamic, overlapping experts inspired by Good's subassembly theory for greater efficiency and continual learning. * directing research into a "Causal Attention Head" designed to implement a form of Good's causal calculus, moving beyond the limitations of standard attention mechanisms. * driving the creation of a new learning paradigm, "Causal Reinforcement Learning from Self-Correction (CRLS)," to replace external reward models with a more robust internal credit assignment process. * championing our primary safety initiative: the "Internal Alignment Governor," an architectural solution inspired by Good's "centrencephalic system" to provide homeostatic regulation and defend against misaligned goals, solving the inner alignment problem.

• provided input and feedback for human reinforcement learning algorithms to enhance foundational and reasoning models for graduate-level science questions. • investigated Python AI/ML developer journey for a major chip manufacturer, optimizing code profiling tools. • constructed deep reinforcement systems for playing board games at a high level.

APImetrics was a Seattle-based startup. It provided the best-in-class active API performance and quality monitoring and governance solution. APImetrics merged with Contxt in late 2023 to form APIContext, which continues to provide the APImetrics solution * developed systems for monitoring APIs utilizing state of the art artificial intelligence/machine learning tools in Python in Jupyter Notebook to generate summary reports from API performance monitoring data using Foundation Models. including OpenAI, Gemini and Claude, for prototypes and PoCs to compare of LLM arithmetic performance * used Docker to make AI and other system services available by API * created PyTorch-based system for unsupervised detection of outliers in time series of API latency data * co-invented CASC (Cloud API Service Consistency) blended quality score system awarded US patent 10644962 * architected and implemented Python system to detect security issues from API specifications for global software provider * developed Python software for automation of regulator-mandated monthly compliance reports for UK banks * researched, generated datasets and wrote white papers on the API/network quality of cloud services, UK banks and leading web services * worked with product team to define new features including cURL importer and UI/UX, and understand end user use cases * provided mentorship to junior colleagues and interns * undertook pre-sales and post-sales engineering including sales support, product evangelism, developing demonstrations and proof-of-concepts, client meetings and providing training * provided technical support to clients including HSBC, Microsoft, NatWest Group, Nokia, Santander, Softbank and USBank * represented APImetrics on industry bodies including as contributing member of the OpenAPI Initiative SLA SIG * provided thought leadership through writing company blogs and being a speaker at public-speaking events including APIstrat, APIdays and API Specifications conferences

* developed, negotiated and executed business agreements for complex, commercial technical consulting and professional services projects for the researchers at Imperial College with primary focus on computer science, data science, machine learning and mathematics * managed relationships with internal stakeholders, including senior academics at Imperial College external clients and partners from startups, SME companies and multinational including C-Level executives * coordinated with clients and researchers to understand project requirements and build strategic partnerships

* project managed the full development cycle of the software including the delivery on time, to budget and quality of the innovative InfoWorks 2D flood prediction product involving - capture of requirements from commercial and technical stakeholders - specification and design of software product - programming in C++ and Fortran under Windows

* wrote technical sections of successful Etisalat bid for Egyptian 3G mobile telecoms license * data modeling in Excel capital/operational expenditure for valuation exercise of a European 3G operator start-up

Consultant creating and applying a strategic framework for European market entry for Extreme Mobile, youth-orientated mobile virtual network operator to be launched summer 2005 (MBA group dissertation project).

Part-time consultant advising on technology transfer strategy to spin-out companies (MBA Assistantship)

* researched telecoms market and industry trends to understand strategic environment of clients * identified and analysed client requirements to create solutions that matched their strategic needs * coordinated the delivery and implementation of these solutions to 3 and Vodafone * wrote technical responses to bid requests from mobile telecoms operators including successful bid response for Vodafone EIR contract * created and delivered technical and commercial presentations to Vodafone, BT, 3 and T-Mobile * undertook competitor analysis of rival vendors * developed feature release timetables with product management teams in Germany and Austria * managed and developed client relationships to influence strategic purchasing decisions * negotiated with senior management as employee representative during internal reorganisation

* troubleshooting technical support issues for mission-critical Unix-based mobile enterprise systems for BTCellnet and Orange

* undertook software development tasks for a realtime data archiving system in C under Digital Unix and VxWorks.

* conducted security-classified computational and experimental research into gas storage and underwater acoustics * wrote three security-classified research reports for the Ministry of Defence client