Seailles Julien's picture

Seailles Julien

Jseailleslighton

AI & ML interests

๐Ÿš€ Passionate about leveraging Large Language Models (LLMs) to unlock new use cases and drive automation, I believe these tools hold immense future value. ๐Ÿ’ก Currently, my work revolves around applying LLMs for auto-generating content from proprietary data. I've developed a method that can analyze any document type and provide comprehensive insights based on user needs. I have developped an agent that can autonomously analyze any content to generate a detailled analysis just based on the data source. ๐Ÿ“Š Additionally, I have developped methods to analyze CSV files using LLMs and emebeddings. ๐Ÿ”ฌ In an effort to enhance model resilience against toxic content, I had the opportunity to conduct red teaming over a large language model. In order to use RLHF to improve this model. ๐Ÿ’ญ To mitigate hallucinations in LLMs, I worked on creating datasets by reverse-engineering their hallucinatory tendencies also in an effort to create RLHF fine-tunning Database. โš™๏ธ Beyond technical applications, my interest lies in addressing strategic challenges companies face with emerging technologies. My goal is to empower them to harness these advancements effectively for maximum value creation. โœ… Past projects include crafting Q&A chatbots using embeddings ๐Ÿค– , automating report writing for a medical company ๐Ÿ’ผ , and conducting audio component content analysis ๐Ÿ”ˆ

Organizations

models

None public yet

datasets

None public yet