Über Oleg
- reproducible LLM and AI evaluation;
- longitudinal comparison across model and prompt changes;
- model, prompt, and workflow drift monitoring;
- validation and controlled revalidation;
- versioned datasets and expected-behavior baselines;
- traceability and audit-ready reporting;
- reliable Python evaluation pipelines for production environments.
- evaluation methodology and scoring design;
- baseline and regression datasets;
- expected-behavior annotation frameworks;
- model and prompt comparison pipelines;
- aggregation, prioritization, and decision rules;
- evaluation evidence and audit-readiness documentation;
- operational validation processes tied to system changes.
- an AI system is moving from experimentation into production;
- model or prompt changes must be compared reliably;
- evaluation results need to remain reproducible over time;
- teams need defensible evidence for governance, risk, compliance, or external review;
- existing evaluation workflows have become fragmented, manual, or difficult to audit.
Englisch
Muttersprachlich oder zweisprachig
Projekt- und Berufserfahrung
- Independent Consultant (IndrasNet/DLX initiative)LLM Evaluation, Validation & AI Governance ArchitectSOFTWARE-HERSTELLERMärz 2022 - Heute (4 Jahre und 5 Monate)Nuremberg, DeutschlandI design and build evaluation, validation, and governance systems for LLM-driven workflows where outputs must remain reproducible, comparable over time, and defensible under technical, risk, or audit review.Current work includes:
- versioned evaluation datasets and expected-behavior baselines;
- reproducible Python evaluation pipelines;
- scoring, aggregation, matching, and prioritization logic;
- regression checks across model, prompt, tooling, and workflow changes;
- traceability between source inputs, evaluation evidence, decisions, and outputs;
- validation and revalidation procedures;
- audit-readiness and governance reporting;
- bounded human-AI review workflows with explicit authority limits.
I also develop reliability controls for evolving AI systems, including drift detection, evaluator consistency, evidence lifecycle management, and release gates that prevent unsupported outputs from being promoted into production or governance processes.The work is designed for high-assurance and regulated environments where reliability, accountability, and historical comparability matter. - Digital Montenegro GmbHCEODIGITALAGENTUREN & IT-CONSULTINGJuni 2018 - Februar 2022 (3 Jahre und 8 Monate)Bar, MontenegroLed digital business operations, financial oversight, partner coordination, and scalable process development across evolving digital initiatives.Key responsibilities included:
- business strategy and operational management;
- budgeting, financial control, and performance monitoring;
- coordination of external partners and stakeholders;
- design of repeatable delivery and reporting processes;
- risk-aware decision-making in changing operational environments.
This experience provides the management, financial, and governance foundation I now apply to AI evaluation, validation, and audit-readiness projects.
Empfehlungen
Sei die erste Person, die Oleg empfiehlt
Teile Deine Erfahrung aus der Zusammenarbeit mit diesem Freelancer.
Diese Freelancer passen auch zu Ihren Kriterien
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Ausbildung und Abschlüsse
- Master’s Degree in FinanceOdessa State Economics University1998Studied Finance and Credit with specialization in public budgeting and securities. The program covered financial systems, capital markets, state budget management, and financial analysis.
- Postgraduate Studies in Financial ManagementOdessa State University2000Focused on business administration and financial management, including organizational finance, operational planning, and financial decision processes.