I build customer-facing AI and operations systems in Vienna. My work sits between discovery and delivery: understand a messy workflow, translate it into clear use cases and acceptance criteria, build a tested prototype or API, and explain the result to technical and non-technical stakeholders.
I am targeting Customer Success in AI, Forward Deployed AI, AI implementation, solutions engineering, and agent-evaluation roles.
- MILLE — turns plain-language ML requests and optional datasets into testable, agent-ready blueprints with deterministic quality gates.
- Jenga — explainable appointment-risk scoring and cancellation-slot recovery for clinics and service businesses.
- AL-MIRAAH — a deployed five-tool MCP system and public dataset for grounded classical- Arabic analysis.
- PHYSIS — a tested hotel-operations feasibility engine for room readiness, staffing, travel time, and maintenance constraints. A sanitized public case study is in preparation.
- The Social Hub, Vienna (2023–2025): guest service and back-office work spanning reservations, check-in/check-out, issue resolution, internal coordination, invoicing, and account reconciliation.
- CareHub project phase (2023–2025): co-developed a telemedicine concept and prototype for digital appointment booking and consultation in Austria, including patient-first UX, scheduling logic, compliance research, and stakeholder communication. It remained a project-stage prototype and is not presented as a registered company or launched product.
Discover the workflow → define the use case → build the system → test failure cases → explain and improve
Python · FastAPI · SQLAlchemy · REST APIs · MCP · Hugging Face · pytest · statistical evaluation