Serkan Yaşar

Serkan Yaşar

AI | ML Developer

Back-End Developer

Data Scientist
📋

Professional Experience

  • Coşkunöz Holding 📍 Bursa, Türkiye

    Software Developer 🗓️ 2022 - Present

    • Developed and deployed AI-powered solutions in industrial settings, focusing on Generative AI, LLMs, Computer Vision, and Machine Learning to optimize production workflows and quality control.
    • Designed and implemented RAG pipelines using Vektor databases for intelligent document retrieval and domain-specific knowledge integration.
    • Designed and deployed LLM orchestration pipelines with LangChain, CrewAI, and n8n, integrating MCP tools to power intelligent chatbots, document Q&A systems, and AI assistants via Streamlit.
    • Applied LoRA and other parameter-efficient fine-tuning techniques for domain-adapted LLM enhancements.
    • Designed and deployed automated quality inspection systems using YOLO, OpenCV, OCR, and PyTorch, reducing manual checks and improving accuracy by up to 95%.
    • Created AI systems that enabled forklift-human collision detection, resulting in a 90% reduction in workplace accidents, awarded in the MESS Occupational Safety Stars Competition.
    • Built operator monitoring systems that prevented faulty products, winning Tofaş Best Kaizen Award, TİSK Digitalization Award, and Ministry of Industry and Technology Productivity Project Award.
    • Developed edge AI systems on Jetson Nano and Raspberry Pi, enabling low-latency inference for real-time industrial scenarios.
    • Implemented microservices with FastAPI, Django, and .NET Core, and managed API integration and deployment using Docker and CI/CD pipelines.
    • Integrated IoT sensor data and PLC communication (e.g., snap7) for fault detection and signal processing in live production environments.
    • Leveraged SQL, NoSQL, and ORM frameworks to manage complex data flows and enable real-time dashboards via Power BI, Tableau, and Grafana.
    • Used RPA (Robotic Process Automation) to streamline document processing workflows, achieving annual savings of €11,000.
    • Conducted large-scale data processing using Apache Spark, Hadoop, and Kafka, with real-time visualization.
    • Experienced in prompt engineering, LLM model evaluation, and staying up to date with multi-agent architectures, hybrid search, and AI trends.
    • Experienced in leading cross-functional teams and collaborating across departments to deliver end-to-end AI solutions in production environments.
🎓

Education

  • Bursa Technical University 📍 Bursa, Türkiye

    BSc in Mechanical Engineering 🗓️ 2018 - 2023

    • Graduated with GPA 2.97. TEKNOFEST participant with HACİWATT team.

  • Istanbul University 📍 Istanbul, Türkiye

    Associate Degree in Computer Programming 🗓️ 2022 - 2024

    • Focus on software development, backend, databases, and algorithms.
    • Expected graduation with GPA 2.8.
🏅

Certifications

MCP Agent Tool Adapter

 


From Protocol to Intelligence: Powering Agents with MCP 

📍 Overview

MCP Agent Tool Adapter enables modular tool invocation via the MCP protocol, and provides agents that can dynamically reason with tools using either Google ADK or LangGraph.

This project transforms MCP tools into:

  • 🤖 Google ADK-based agents with streaming FastAPI or CLI interfaces
  • 🧠 LangGraph-based agents that use ReAct + streaming tool planning

Project Link: https://github.com/serkanyasr/mcp-agent-tool-adapter

Features:

  • 🚀 Converts MCP tools into LangGraph and ADK agents

  • 🧠 Agent collaboration and task orchestration

  • 🔧 MCPClient/MCPServer-based modular backend

  • 🖥️ CLI chatbot interface powered by OpenAI

  • 📁 Supports multiple data backends: memory, SQLite, and filesystem

  • 🛠 Configurable through mcp_config.json