Životopis
Obrazovanje
Harvard University
A.B. Računarstvo; sekundarna oblast neuronauka. Maj 2027.
Predmeti uključuju računanje na skali, reinforcement learning, AI safety, biološku i vještačku inteligenciju, operativne sisteme, algoritme, mašinsko učenje, vjerovatnoću, nauku o podacima, hardver i računarsku neuronauku.
Iskustvo
Refractal
Forward Deployed Engineer, 2026. Building the runtime security layer for production AI agents: least-privilege tool access, sandboxed execution, permission boundaries, approval gates, and auditable control over agent actions.
Office of the Prime Minister, Montenegro
AI Engineer & Technology Advisor, 2026. Designed and built government AI systems spanning internal knowledge, legislation, public records, retrieval, and agentic workflows, while advising on sovereign AI architecture, compute infrastructure, and national technology strategy.
Zenlytic
Software Engineer Intern, 2025. Built evaluation, fine-tuning, and continual-learning systems for an AI analytics agent, including a closed-loop pipeline from verified answers into training data.
Coinis
Software Engineer Intern, 2024. Built ad infrastructure and early AI systems for generating and ranking high-converting landing pages.
Projekti i istraživanje
MNE-MCP
A machine-readable national data layer across 27 government sources and 1M+ raw records, with MCP tools for source-grounded access to legislation, companies, procurement, courts, finance, and public records.
Sovereign BCMS LLM
Training open models for government use across BCMS language, legislation, and public-sector knowledge.
Talas
A people-intelligence system for finding the right people across the Balkans and its global diaspora.
Voice MNE-MCP
A voice interface for querying Montenegro’s legislation and public records naturally.
Harvard Rover
Software team lead for an autonomous rover combining perception, SLAM, planning, and navigation. 2024–present.