• Conduct literature reviews on LLM-related topics (Federated LLM, Data Privacy, Security, etc.).• Assist with ongoing research projects at SAIL.• Participate in team meetings, reading groups, and brainstorming sessions.• Assist in preparing research publications, reports, and presentations based on the research findings.

• Teaching Assistant at SkillPixel, an education startup focused on AI.• Developed and assignment Python notebooks to support student learning and engagement.• Assisted with course material preparation and provided guidance to students on AI-related topics.• Collaborated with instructors to design assignments aligned with course objectives.

• Developed Python and Bash scripts using LLM inference libraries (TensorRT-LLM, Aphrodite Engine, DeepSpeed, LLM-Swarm, TGI, vLLM) to distribute batch inference workloads across multiple GPU nodes, reducing computation time.• Benchmarked large language models (Mixtral, Llama 2) under various conditions to optimize performance.• Conducted experiments on large datasets (Spider, BioASQ) to test and improve inference throughput.• Created documentation and presented a tutorial on the… Show more • Developed Python and Bash scripts using LLM inference libraries (TensorRT-LLM, Aphrodite Engine, DeepSpeed, LLM-Swarm, TGI, vLLM) to distribute batch inference workloads across multiple GPU nodes, reducing computation time.• Benchmarked large language models (Mixtral, Llama 2) under various conditions to optimize performance.• Conducted experiments on large datasets (Spider, BioASQ) to test and improve inference throughput.• Created documentation and presented a tutorial on the distributed batch inference scripts, showing significant efficiency gains. Show less