At Cogerphere AI Labs, I build data-first academic language models and production AI systems for LaTeX correction, formal writing, code, and research assistance.
I'm a core researcher and co-author of Meridian 0.1 and Meridian-AC-Nano, curating instruction datasets, fine-tuning models with LoRA, and evaluating training, safety, reasoning, quantization, and GGUF deployment. Our measured ROUGE-L scores reached 0.641 for Meridian 0.1 and 0.929 for Nano.
I also contribute to OpenBentt and build agentic research workflows, RAG pipelines, AI APIs, and backend microservices using Node.js, PostgreSQL, MongoDB, AWS, and WebSockets.
