I am a hands-on engineer with 7+ years of deep engagement in Machine Learning (from early seq2seq/GANs to modern Agentic AI systems) and 4+ years of production Full-Stack experience. In my current work, I aim to bridge the gap between "what AI can do" in theory and "what AI should do for us" in practice.
My motivation goes beyond code. I view Artificial Intelligence not as a product for the few, but as a collective achievement that should function as a universal utility. I believe the benefits of this technology belong to the world, and I am dedicated to engineering the infrastructure that turns this potential into a safe, borderless tool that fundamentally improves the condition of life.
Currently, I am focused on building Agentic AIs, Reasoning LLMs while keeping an eye on cutting edge AI/ML research being made by the greatest minds. I apply rigorous engineering standards to orchestrate intelligent workflows that democratize access to intelligence and solve complex problems for the wellness of all living beings.
A web-based app for building and visualizing fully-connected neural networks using TensorFlow.js and p5.js. Users can experiment with tasks like logistic and linear regression while gaining insights into the learning processes of neural networks.
A hands-on implementation of the encoder-decoder architecture with attention mechanisms in TensorFlow, mirroring the foundational ideas behind modern LLMs. This project shaped my understanding of tokenization, attention, and autoencoders.