Enes Gokce

The project was for an enterprise market research company client. I was responsible for developing a retrieval augmented generation (RAG) solution to analyze their data on the R&D side. Once I executed my solution on the R&D side, I worked with the SWE team to put it into production.

Some steps I completed in this project:
• Converted tabular data to textual data with text augmentation.
• Created a RAG system by using LangChain and Llamaindex frameworks
• Tested different implementation ideas that give the best results for this specific client.
• Evaluated LLM results using the LLM Evaluation Benchmark Rubric developed by the in-house data science team (human evaluation).

A hybrid LLM-based retrieval augmented generation (RAG) system that allows users to extract insight from big survey datasets. With this system, the user can communicate and explore survey data better.

Tools: PostgreSQL vector database, Bedrock API, Claude 3 model, Mistral 7B model, word embeddings

It's a project that generates interview questions based on job descriptions posted on a job board website. My AI solution generated five interview questions from different categories for each job role category.

Tools: Amazon Bedrock, Amazon SageMaker, Claude 3, prompt engineering

Worked on topic generation for each document.
• Did literature review.
• Created R&D roadmap based on the literature review.
• Created a demo output.
• Communicated with shareholders about the feature development process.
• Delivered the R&D part of the solution for topic extraction and topic classification.
• Worked with the SWE team on the project deployment.

• Developed novel and accurate NLP systems using generative AI and large language models (LLMs).
• Contributed to named entity recognition (NER), text summarization, and emotion classification systems.
• Prepared and presented reports on the NLP AI engine for investors and client onboarding.
• Developed the R&D part of the Pinecone vector database solution for the RAG conversational AI chatbot system.
• Monitored NLP repositories' logs on AWS Cloud Monitor to ensure optimal performance of AI algorithms.
• Collaborated closely with the product team, updated them, and prepared technical documentation.
• Led a team to build a chatbot system using the retrieval augmented generation (RAG) framework.