BeLedgerReady
An audit-readiness assistant for SMEs that combines deterministic financial tests, explainable anomaly analysis, evidence tracking, and transparent reporting without presenting itself as an auditor.
View repository →AI & data education · applied AI systems · computational research
Building AI systems for discovery, learning, and consequential decisions.
AI and data educator, mentor, certification juror, and independent researcher working across applied AI systems, computational research, analytics, and technical education.
My work connects applied AI systems, scientific exploration, analytics, and pedagogy. I am particularly interested in systems that make evidence, uncertainty, and reasoning visible while keeping human judgement firmly in the loop.
What I build
A portfolio shaped by three connected practices: applied AI systems, computational research, and technical education.
Applied & Decision-Support AI
Auditable AI systems, anomaly triage, agentic workflows, and human-in-the-loop tools for situations where mistakes matter.
Computational Research
Scientific data pipelines, signal and time-series analysis, anomaly detection, multimodal exploration, and uncertainty-aware prototypes.
Technical Education
Higher-education teaching, agentic learning tools, curriculum design, mentoring, and project-based paths from theory to implementation.
Technical formation
A sustained technical development path, not a recent pivot into AI.
My technical practice has developed continuously since 2017 through structured study, professional programmes, self-directed learning, and project work across software development, machine learning, deep learning, reinforcement learning, autonomous systems, computer vision, cloud technologies, data science, and generative AI.
This included learning through ecosystems such as Udacity, AWS, Coursera, Intel/OpenVINO, specialist AI academies, and other professional technical programmes, culminating in an MSc in Artificial Intelligence with First Class Honours in 2026.
The projects on this site are part of that longer trajectory: they document an evolving technical practice in which learning, building, research, explanation, and evaluation reinforce one another.
Selected projects
Current and representative work across auditable AI systems, public-interest technology, scientific exploration, agentic architectures, and multimodal creation.
Airline Said No
An explainable passenger-rights assistant that structures a disrupted-travel case, evaluates the available evidence, and helps travellers understand plausible rights and next steps while making uncertainty explicit.
Recent projectIndustry Integrated AI Systems Synthesis
An integrated, governed AI architecture for anomaly triage in safety-critical aerospace operations, combining multiple analytical components with escalation logic, uncertainty handling, and human authority.
View repository →PyroNav
An offline-first wildfire evacuation PWA designed around hazard-aware route diversity, combining fire observations, weather, terrain, uncertainty, and official closures without replacing emergency instructions.
In developmentintibi
A desktop solar companion whose evolving states are driven by daily H-alpha observations. The project translates real solar behaviour into an interactive, scientifically grounded character system.
In developmentBioacoustic Topology
An exploratory framework that models birdsong as a latent dynamical system through acoustic features, manifold embeddings, motif grammars, trajectory reconstruction, and transition analysis.
View repository →Latent Physiological Topology
An exploratory, non-diagnostic investigation of cardiac acoustic signals as latent physiological state spaces, using manifold learning, trajectory dynamics, cyclic representations, and anomaly analysis.
View repository →Design of Agentic Workflows
A systems-oriented study of multi-agent workflow design, with explicit roles, handoffs, validation stages, escalation paths, and governance mechanisms for more dependable orchestration.
View repository →Moonlit Showrunner
A multimodal short-film pipeline that turns a story premise into a structured production workflow spanning narrative design, storyboards, character references, generated imagery, animatics, optional video clips, and final assembly.
Research direction
Independent, project-based research on AI systems for scientific triage, uncertainty-aware exploration, and human attention.
I explore how AI can help humans notice meaningful signals in complex scientific and technical environments: surfacing anomalies, connecting multimodal evidence, tracking uncertainty, and supporting disciplined investigation.
Current themes include scientific triage, AI mission control, human–AI discovery cockpits, agentic technical education, and multimodal frontier detection. The work is documented through reproducible repositories, technical specifications, notebooks, and research-oriented prototypes.
Teaching, mentoring, and communication
Technical education is both a professional practice and a test of whether systems and ideas are genuinely understandable.
- Higher-education teaching and curriculum design across AI, data, cloud, business analysis, and digital systems.
- Professional mentoring in data analysis, business intelligence, machine learning, Python, SQL, and applied AI projects.
- Project assessment and certification-jury work, with emphasis on technical reasoning, evidence, limitations, and responsible use.
- Learning systems, visual resources, simulations, and agentic teaching tools that help learners reason rather than merely reproduce procedures.
Current interests
Questions and systems I am actively developing.
AI mission control
Interfaces and governed agent systems that help humans coordinate evidence, uncertainty, priorities, and intervention.
Frontier detection
Multimodal systems for finding unusual, weak, or easily overlooked signals across scientific data and literature.
Agentic technical education
Learning environments in which AI agents support investigation, feedback, simulation, and skill verification without replacing learner judgement.