Selma Kocabıyık

Junior AI Developer

• Built The Chunker — transforms Floorplanner's design database into structured formats optimized for LLM and RAG systems.
• Experimented with multiple chunking strategies to improve vector search performance and SQL-based querying.
• Built a RAG pipeline using LangChain and Qdrant for semantic retrieval over large design datasets.
• Built a manifest generation system for clear data lineage and easier debugging.
• Explored MCP-based integration patterns and quality-control automation within the design data pipeline.

AI Engineer Intern

• Designed and deployed AI-integrated automation workflows using Claude AI and Model Context Protocol (MCP).
• Optimized context-aware data exchange across Google Sheets, Apollo, Instancy, and custom APIs.
• Built and maintained automation systems that reduced manual workload and operational costs.
• Managed high-volume API integrations, real-time triggers, and prompt-tuned tool chaining in low-code workflows.

Data Scientist

• Designed and implemented Python-based web scraping systems for Amazon and Alibaba product data.
• Cleaned, normalized, and preprocessed datasets for reliable cross-platform comparison.
• Built a price-comparison pipeline to identify lower-cost sourcing opportunities.
• Generated actionable insights to support cost optimization and sourcing decisions.

BSc Artificial Intelligence

Thesis: Enhancing Workflow Automation with MCP and Claude AI

Exchange — Computer & Information Sciences

Pre-University Education (VWO)