-Developed a Q&A chatbot on AWS using API application, leveraging LLMs. Implemented fine-tuning with QLoRA technique for improved performance, alongside web scraping and RAG for retrieving Ericsson website data. Used prompt engineering and parameter optimization to reduce prompt response time from 20s to 2s significantly. - Tools: Chromadb, LangChain, FastAPI, AWS, Docker, Whisper, SDK- LLM-based replicator for 3D objects. Implemented a system empowered by a similarity… Show more -Developed a Q&A chatbot on AWS using API application, leveraging LLMs. Implemented fine-tuning with QLoRA technique for improved performance, alongside web scraping and RAG for retrieving Ericsson website data. Used prompt engineering and parameter optimization to reduce prompt response time from 20s to 2s significantly. - Tools: Chromadb, LangChain, FastAPI, AWS, Docker, Whisper, SDK- LLM-based replicator for 3D objects. Implemented a system empowered by a similarity score algorithm capable of replicating 3D objects based on voice input sourced from a repository. -Tools: Hugging Face, Unity, C#, Whisper Show less

Created a recommendation system that analyzes customer purchase history to identify similar buying patterns. The system suggests interesting products to customers based on past purchases using association rules. Tools: Numpy, Pandas, scikit-learn, Surprise,Matplotlib , SQL- The Text-Based Emotion Detection Algorithm was developed to identify emotions within email or chat text. The model achieved an accuracy rate of approximately 87% by employing supervised training on the… Show more Created a recommendation system that analyzes customer purchase history to identify similar buying patterns. The system suggests interesting products to customers based on past purchases using association rules. Tools: Numpy, Pandas, scikit-learn, Surprise,Matplotlib , SQL- The Text-Based Emotion Detection Algorithm was developed to identify emotions within email or chat text. The model achieved an accuracy rate of approximately 87% by employing supervised training on the company dataset. Tools: NLTK, scikit-learn,Numpy, Pandas, Keras - I utilized tools such as Numpy, Pandas, SciPy, Matplotlib, Seaborn, Streamlit, and PowerBI for data analysis, preparation, and visualization. Show less