Can we trust what AI does?
Beyond prompt tweaks and wrapper APIs lies real architecture: ML validation, dynamic model routing, and production reliability. My trajectory is a deliberate chain built to answer that single question.
Engineering AI Systems with Empirical Rigor
Hi, I'm an AI & ML Developer. My focus lies at the intersection of machine learning systems, prompt routing, and production software architecture. Rather than relying on black-box API wrappers, I construct transparent pipelines that evaluate, route, and deliver reliable AI capabilities.
Each step forced the next
Pivot
Walked away from surface-level abstractions. Committed to learning how systems actually work under the hood.
EARLY
Built ML evaluation pipelines from scratch. Learned that intuition without metrics is just guessing.
read case study →Router
Extended model selection into dynamic LLM routing. One model doesn't fit all tasks — so I built the system that picks.
read case study →Gesenu
Synthesized everything into a shippable product. The real test: can a real user actually use this?
read case study →Project Case Studies
EARLY
Validated machine learning fundamentals, feature engineering, and rigorous model evaluation pipelines.
Router
LLM & Agent engineering system for dynamic model routing, prompt optimization, and fallback orchestration.
Gesenu
Full-stack product delivery, bringing complex agentic workflows into an intuitive, polished user experience.