I help engineers moving into AI research, students targeting top-tier conferences (NeurIPS/CVPR/ECCV/ICML), and academics moving into industry — get unstuck faster than they would alone. What you actually leave a session with: a sharper problem formulation, an experimental design that survives review, or a concrete fix to your paper's narrative. Not generic advice — the same lens I use on my own submissions. Currently an AI Research Engineer at AMD working on vision-language-action and world-action models and inference optimization (up to 30x speedups). I co-authored DUET-VLM (CVPR 2026), authored Pix2Gif (ECCV 2024), and have run LLM and agent fine-tuning workshops at IIT Bombay and IISc plus GenAI hackathons with 200+ participants. I mostly mentor on research and AI guidance: research problem formulation and experiments, paper writing and rebuttals, breaking into ML research or applied AI roles, and hands-on AI such as fine-tuning LLMs/VLMs and making models faster. If you're serious about building depth, not just stacking models, we can work.