Aco Kovaceski
I turn large, messy information into something structured enough to act on, and I build the systems that do it.
I have had congenital nystagmus all my life: my eyes move on their own, and text slides when I read. So I learnt early to read everything, structure it, and check it against the source. That habit became a career. I founded Wizard Computer Training in 1990, then worked at IBM Australia, and have spent the decades since designing data and analytics platforms for insurance, government, defence and industry. I run an automated trading fund, and I hold a doctorate in leadership and decision making, on cognitive bias in leaders.
Selected work
- Commons A local-first research system for doctoral students. It keeps every claim tied to its source, reads the new literature each morning, and looks for the evidence against your argument as hard as the evidence for it. Bring your own model.
- Philotas A real-time geospatial common operating picture: many live sources fused onto one map and connected by an AI-driven ontology, with a knowledge graph, time replay and classification-based access control.
- Modelwrite The model you can prove. Open source analytics for SysML: it reads the models you already have, measures coverage and traceability baseline by baseline, and answers questions with the evidence attached.
- Bylazora The migration you can prove. A byte-exact engine that checks a migrated mainframe workload's outputs against its legacy COBOL reference.
Writing
- The Protective Expert: knowledge-guarding in teams Why organisations lose more to guarded knowledge than to missing knowledge, how deep an intervention has to go to shift it, and the mirror-image failure of teams that always think the grass is greener.
- The whole point of MBSE was the model everyone could see What four teams building four different products taught me about why systems engineers write things down.
- Creativity is a mode, not a gift Five conditions for the open mode, and why most managers practise the opposite.
- The great AI pricing paradox Benchmark quality per dollar across twenty-odd models, and why you are probably overpaying for language models.
- The best nystagmus drugs might already be on the shelf I pointed my fund's tooling at my own eyes: 3,200 papers, 669 findings, and a list of cheap off-patent drugs nobody had trialled for nystagmus.
Contact
enable JavaScript to see the addressWrite