Igor Nesterenko
Igor Nesterenko is a Bristol-based systems analyst and IT delivery manager. He makes systems legible, then ships them — and he writes here about the judgement that has to stay human when AI is in the loop.
Who is Igor Nesterenko?
Igor Nesterenko is a systems analyst and delivery manager based in Bristol, United Kingdom, with the right to work in the UK. For more than eight years he has turned ambiguous business goals into specified, delivered software across e-commerce, logistics, SaaS, CRM and ERP — often inheriting stalled products and getting them in front of users.
The through-line is judgement under complexity: deciding what to build, what to verify, and who answers for the result. Recent proof points include 30+ security audits, a reconciliation of more than five million inventory records, and an ERP integration spanning custom software, Microsoft Dynamics 365 and Sage 200. On one logistics platform he led a team of seven to a working MVP in six months after a year of prior delay.
He is the UK point of contact for Qorym, an AI-assisted delivery practice; collaborates with Secureware on security audits and ERP integration; and founded Salience, an agency focused on generative engine optimisation (GEO) and answer-engine visibility. He writes the Opinion & Analysis essays on this site. The argument that runs through them is practical: AI is most useful when a human still owns the verification.
What AI research does he do?
Alongside delivery work, Igor pursues hands-on research in applied machine learning and AI engineering, currently while completing a part-time MSc in Ethical Hacking and Cybersecurity at Abertay University. Three strands:
- Machine learning for network security. Research into ML approaches for detecting and filtering malicious data in network traffic — anomaly and intrusion detection that separates hostile patterns from ordinary noise, with an emphasis on methods that hold up outside the training set.
- AI-assisted software development. Practical study of how AI coding tools actually change delivery — where they accelerate work, and where they shift effort from writing code to verifying it — so teams can adopt them without quietly importing risk. Day to day this is a production practice: Cursor, Claude Code, and automated QA in an 80/20 prompt-driven pipeline.
- Retrieval-augmented generation (RAG). Building RAG systems that ground large-language-model output in verified sources, so answers can be traced back to evidence rather than taken on trust — the engineering counterpart to the critical-thinking themes of this site.
Selected focus areas
Background
Igor holds an Undergraduate Advanced Diploma with Honours (Level 6) in IT Systems Analysis & Design from the University of Oxford, and is completing a part-time MSc in Ethical Hacking and Cybersecurity at Abertay University (Year 1 completed; Securi-Tay 2026). He volunteers technical skills with Oxfam and an environmental charity in the South West. His name is also transliterated Ihor Nesterenko.
Full professional history, project portfolio, and CV: nesterenko.co.uk.
Essays on this site
- Data & cognitionModel Collapse and Digital Dementia: Causation, Correlation, or a Convenient Rhyme?
- DiscoveryAI-Assisted Discovery Audit Before Workshop Sign-Off
- TrustShould I Trust AI?
- InterpretabilityA Chain of Thought Is Not a Train of Thought
- StrategyAI Strategy Is Only Precise Where You Can Be Proven Wrong
- Data & trainingThe AI Tautology Crisis Is a Data-Provenance Failure
- EducationGenAI Skill Erosion Is a Governance Problem, Not a Technology Verdict
- Labour & AIRehiring the Engineers AI Was Meant to Replace
- EducationUniversities Guarantee Critical Thinking — If You Design for It