Portrait of Vladimir Stantchev

Vladimir Stantchev

Professor of Business Informatics · Managing Director of the Institute of Information Systems, SRH University of Applied Sciences Heidelberg, Campus Berlin
Principal Engineer, PRESIDIO Group, Sofia

I am Professor of Business Informatics at SRH University of Applied Sciences Heidelberg and Principal Engineer at PRESIDIO Group, and I am based in Sofia. My research is computational jurisprudence: law and policy treated as properties that a computation must satisfy while it runs.

A rule that cannot be checked at machine speed does not bind a machine. I work on making it checkable: authority that AI agents delegate to each other through capabilities that can only shrink, and evidence of every consequential action that a party who does not trust the agent can verify. The software behind the papers is open source, with signed releases and archived DOIs.

Google Scholar: 2,774 citations · h-index 28 · i10-index 55 (as of 22 November 2026)

ORCIDGoogle ScholarOpenAlexSemantic ScholarGitHubPRESIDIO on GitHub

Get in touch

Research: computational jurisprudence, AI governance and agentic AI security

Computational jurisprudence asks what it takes for a norm to bind software that acts on its own. The answer I pursue is runtime compliance: the obligation is compiled into the system, and the system produces evidence that it was met. Four lines of work carry that programme.

Capability-based security for AI agents

Delegated authority that can only be narrowed as it passes from agent to agent, and that still holds when the issuer is offline. Capability-based security gives the agentic AI security model its spine.

Verifiable computation and evidence

Signed, parent-linked evidence records and ledger-anchored audit logs, so that a third party can check what a system did without trusting its operator.

AI governance and IT governance

The IKI-Gov reference model (Springer, 2026/2027) takes classical IT governance to AI governance, with lifecycle gates and measurement dimensions. It also covers audits of AI-supported decision systems: does the optimiser earn its place?

Privacy engineering and information security

Privacy engineering for licensed AI training data and for agentic payments, where metadata is filtered before a payment executes, and information security management in practice.

The question is older than the agents. Earlier I asked it of distributed systems, in architectural translucency and service-oriented computing: what can an operator verify instead of trust?

Topics: Computational Jurisprudence · AI Governance · Agentic AI Security · Runtime Compliance · Capability-Based Security · Verifiable Computation · Privacy Engineering · Information Security · IT Governance · Service-Oriented Computing

Selected publications

All 121 publications · Also as BibTeX and CSL-JSON

Research software and projects

Every result in the recent papers can be rerun from the code behind it. That code is published as PRESIDIO: 45 open-source repositories under github.com/presidio-v, 15 packages on PyPI published with trusted publishing and attestations, signed release tags, and Zenodo DOIs for the research artefacts.

Funded projects

  • ITG4TU, IT Governance for Tunisian Universities (European Union)
  • ITG4AU, IT Governance for Albanian Universities (European Union)
  • OpSIT (German Federal Ministry of Education and Research (BMBF))
  • CISSOS, with the municipal administration of Getafe, Madrid (European Union)

Sofia: research collaborations and the Bulgarian startup ecosystem

Sofia, Bulgaria, is my base. I work there as Principal Engineer at PRESIDIO Group, and I hold my professorship at SRH University of Applied Sciences Heidelberg, Campus Berlin.

Research collaborations

Sofia and the Bulgarian startup ecosystem

I care about this ecosystem, and much of my work outside the university goes into it. The numbers show why it deserves the attention. The Sofia region, Yugozapaden, "ranks first among all EU-27 regions in the share of employed ICT specialists", and "roughly 86 percent of the country's startups are located in Sofia" (Stantchev and Sedlarski, BEP 03‑2026).

Talent at that density is the endowment. What young companies still lack is the trust that lets them sell to a regulated customer abroad, and compliance that a machine can check is one way to supply it. The Burgas paper asks how a tech ecosystem outside the capital closes that gap; in Sofia I work on the same question with early-stage companies, in practice. I offer:

  • architecture and security reviews for early products
  • advice on compliance by design for regulated products, including the EU AI Act, DORA and NIS2
  • mentoring for founders who sell to regulated customers in the EU
  • contact between founders and university research: thesis topics, student projects and joint proposals
  • talks on AI governance for founder and student groups, on request

I also support Sofia Tech Park, the first science and technology park in Bulgaria, where research groups, universities and young companies work on one campus. A park only pays off when what it hosts meets: the laboratory, the founder and the customer. In October 2018 I spoke there on a panel on healthcare and pharma digitalization for patient empowerment, at CEE Block Sofia 2018.

Vladimir Stantchev on stage with three other panellists at CEE Block Sofia 2018; the screen shows his name card for the panel on healthcare and pharma digitalization
Panel "Healthcare and Pharma Digitalization for Patient's Empowerment", CEE Block Sofia 2018, Sofia Tech Park, 26 October 2018.

Startup currently supported: TOVR (Sofia, founded 2026), which describes itself as an AI execution layer for European road freight.

Contact

Email: vladimir@stantchev.com
Profiles: ORCID · Google Scholar · OpenAlex · Semantic Scholar · GitHub · PRESIDIO on GitHub