Specialized in LLMs, RAG applications, and innovative AI solutions at appliedAI Initiative GmbH.
As an AI Software Engineer with APPLIED AI INITIATIVE, my work revolves around developing innovative AI solutions, like a local RAG application tailored for complex data analysis. Collaborating with cross-functional teams, I've been instrumental in deploying this technology using Docker, ensuring scalability across environments.
Currently, as a BSc AI student at Deggendorf Institute of Technology, my academic focus in Artificial Intelligence complements my professional role. My expertise spans across LLMs, RAG applications, and developing AI solutions that address real-world challenges.
For a university project, I developed an advanced post-processing method to enhance deep neural network predictions, specifically addressing label error detection in connected components within semantic segmentation tasks.
Excited to share that I'll be joining appliedAI Initiative GmbH as a Level 2 AI Engineer – LLMs (Working Student).
Grateful for the opportunity and excited to work with Mingyang Ma and the amazing team on cutting-edge NLP and LLM projects.
One of the biggest challenges with large language models (LLMs) is their inability to maintain long-term memory. Typically, models like ChatGPT need constant reminders about previous interactions once they reach their token limit, often leading to random answers or hallucinations.
To address this issue, we can implement a small model that processes user prompts and extracts valuable information to create vector embeddings. This model continually updates the long-term memory of the agent by storing these embeddings in a vector database.
I'm always open to discussing new projects, creative ideas, or opportunities to be part of your vision.