I build trustworthy AI systems, developer platforms, and cloud-security tools— with an emphasis on clear evidence, explicit permissions, and human control.
I am a Technical Support Specialist growing into Cloud, Security, and AI Engineering. My work explores how humans and AI agents can collaborate safely in real products—not only generate answers.
My current focus is Sparkbase and its open-source foundation, SparkKit: infrastructure for multiplayer AI applications where every agent action is governed by explicit permissions, approval boundaries, and audit trails.
tickets.read ALLOW
knowledge.search ALLOW
drafts.create ALLOW
refund.execute APPROVAL
customers.delete DENY
The goal is simple: make powerful AI useful inside real workflows while keeping people in control.
An open-source foundation for building portable software and multiplayer AI applications. It combines a modern application stack with tenant isolation, authentication, organizations, and a path toward policy-controlled agent actions.
TypeScript · Next.js · PostgreSQL · Prisma · Docker
A source-backed research data foundation for satellite and ground-segment intelligence, designed around traceable evidence and explicit knowledge gaps.
TypeScript · Research Data · Evidence Provenance
A bounded research agent that validates what sources can support, preserves uncertainty, and avoids inventing missing answers.
AI Agents · Source Validation · Responsible AI
An identity-visualization concept for Microsoft Entra ID that maps access paths and uses agentic AI to explain issues, detect patterns, and recommend next steps.
Identity Security · Microsoft Entra ID · Agentic AI
- Human–AI collaboration and permission-aware agents
- Cloud architecture, identity, and security engineering
- Evidence-grounded AI and auditable automation
- Open-source developer tooling and portable application foundations
- TypeScript, Python, Next.js, PostgreSQL, Docker, AWS, and Azure
Evidence over confident guesses.
Explicit permissions over invisible autonomy.
Human approval for consequential actions.
Portable foundations over unnecessary lock-in.
I am interested in open-source collaboration, early-stage AI infrastructure, cloud security, and thoughtful conversations about safe human–agent systems.