A researcher building
your business tools.

I grew up in Malkhanagar, Munshiganj, a small town in Bangladesh with limited access to world-class education and almost no exposure to the global AI research community. I started university at AIUB in Dhaka in 2022 and became obsessed with one question: why do AI systems fail when they should work?

That question led me to publish peer-reviewed research at Springer and IEEE, develop the RAGCare-QA medical AI benchmark, and run controlled empirical studies that revealed something surprising, standard RAG pipelines, the most common AI grounding technique, barely outperform random chance in medical question answering. Retrieval quality was the primary bottleneck. Not model size. Not architecture. Just bad retrieval.

That finding became the foundation of CortexLabs. If I could identify exactly why AI fails at a research level, I could build AI systems for businesses that are engineered to avoid those failures. Not assembled from templates. Not configured from no-code dashboards. Actually engineered, with reliability as the first design principle, not an afterthought.

In February 2027 I will begin a Masters by Research in AI at an Australian university, formally affiliating CortexLabs with an international research institution. The lab and the agency grow together, every research finding informs how we build for clients, and every client system surfaces new research questions.

My mission is straightforward: make enterprise-grade AI reliability accessible to every US small and mid-sized business that deserves it.