Ylli Prifti
Ylli Prifti
About
Detail
VP of Product Engineering
Florida, United States
I've spent my career at the intersection of platform engineering and emerging technology—first building systems that needed to earn user trust at scale, now focused on what AI means for platform engineering at scale.
At Anaplan, I lead teams working on identity, authorization, and AI enablement for a planning platform used by millions. The problems I find most interesting sit at the boundaries: How do you build authorization systems that handle 250M+ daily decisions while staying auditable? How do you introduce LLM capabilities into enterprise workflows without creating new categories of risk?
My Ph.D. research focused on computational trust—how trust emerges in online interactions. That lens still shapes how I think about AI systems: reliability isn't just uptime, it's predictability, explainability, and knowing when to defer to humans.
Lately I've been:
→ Building open-source tools for AI-assisted configuration and web data extraction
→ Designing multi-model LLM architectures for enterprise use cases
→ Writing about what works (and what doesn't) at ylli.prifti.us
I share my learnings here—not polished thought leadership, but notes from the workshop floor. If you're working on similar problems, I'd enjoy connecting.
At Anaplan, I lead teams working on identity, authorization, and AI enablement for a planning platform used by millions. The problems I find most interesting sit at the boundaries: How do you build authorization systems that handle 250M+ daily decisions while staying auditable? How do you introduce LLM capabilities into enterprise workflows without creating new categories of risk?
My Ph.D. research focused on computational trust—how trust emerges in online interactions. That lens still shapes how I think about AI systems: reliability isn't just uptime, it's predictability, explainability, and knowing when to defer to humans.
Lately I've been:
→ Building open-source tools for AI-assisted configuration and web data extraction
→ Designing multi-model LLM architectures for enterprise use cases
→ Writing about what works (and what doesn't) at ylli.prifti.us
I share my learnings here—not polished thought leadership, but notes from the workshop floor. If you're working on similar problems, I'd enjoy connecting.
