About

Erol Sen

I'm an AI Engineer based in Izmir, Turkey. I build production-grade agentic systems, the kind that have to be right, recover when they're wrong, and keep their cost and latency under control.

Most of my work lives on LangGraph: multi-agent orchestration, structured memory, and context engineering for long-running workflows. My retrieval stack pairs vector RAG with Graph RAG over Neo4j and deterministic rules for the paths where correctness isn't optional. I ship streaming FastAPI services on Azure, AWS, and GCP, instrumented with Langfuse and validated by LLM-as-Judge grading and schema-first outputs.

Before agents, I spent a few years in low-latency edge computer vision: YOLO, OpenCV, and VLM pipelines on NVIDIA Jetson, which is still where I reach when a problem has to run in real time at the edge.

Experience

  1. AI Engineer · Topraq

    Apr 2026 – Present

    Izmir, Turkey

    Tech & keywords: Agentic AI · LangGraph · CrewAI · LangChain · Multi-Agent Systems · Tool / Function Calling · Graph RAG · Neo4j · Memory Engineering (episodic / semantic / procedural) · Context Engineering (compaction) · LangGraph Checkpointer (Postgres) · Multi-Provider LLM Orchestration · Langfuse · Structured Output · Pydantic

  2. AI Engineer · VIZIO AI

    Sep 2025 – Apr 2026

    Istanbul, Turkey (Remote)

    Tech & keywords: LLM Tool Calling · Conversational AI · Context-aware Image Retrieval · Multi-Agent Systems · LangChain · OpenRouter · Ollama · LLaMA · RAG · GenAI · VLM · Multimodal AI · Computer Vision · FastAPI · Python

  3. AI & Computer Vision Engineer · Tundra AI

    Jul 2024 – Sep 2025

    Prague, Czechia (Remote)

    Tech & keywords: PyTorch · Ultralytics YOLO · OpenCV · DBSCAN · HDBSCAN · NumPy · TensorRT · Jetson · Edge AI · FastAPI · Google Cloud · PyQt5 · Docker · K-Means · Real-Time Measurement · Industrial Automation

  4. AI Engineer · Çağdaş Cam

    Dec 2023 – Jul 2024

    Aydın, Turkey

    Tech & keywords: AI Engineering · Computer Vision · YOLO · Real-Time Inspection · Industrial Automation · On-Prem Deployment · NVIDIA Jetson

Skills

GenAI / Agentic AI

Agentic AILLM agentsMulti-agent systemsLangGraphCrewAILangChainTool / function callingStructured outputsContext engineeringMemory engineeringMulti-provider orchestration (OpenAI, Claude, Gemini)

RAG & Knowledge Graphs

RAGGraph RAGEmbeddings & vector DBsNeo4j / CypherNetworkXHDBSCANLLM-as-Judge evaluationGuardrails

Computer Vision

VLM pipelinesYOLOOpenCVPyTorchTensorFlowTensorRTEdge AI (NVIDIA Jetson)

APIs & Infrastructure

FastAPISSE / token streamingasyncioDockerPostgreSQL / SupabaseAzure · AWS · GCPLangfuse / LangSmithOpenTelemetrypytest

Education

  • B.Sc., Mechanical Engineering

    2016 – 2023

    Afyon Kocatepe University · capstone: automated fruit sorting (Python / OpenCV / ML).

  • ERASMUS+ Exchange

    2020 – 2021

    Universität Duisburg–Essen, Germany · Control Engineering.

Certificates

  • Agentic AIDeepLearning.AI · 2026
  • Long-Term Agentic Memory with LangGraphDeepLearning.AI · 2026
  • AI Agents in LangGraphDeepLearning.AI · 2026
  • Machine Learning Engineer NanodegreeUdacity · 2021
  • Convolutional Neural Networks in TensorFlowCoursera · 2019
  • Natural Language Processing in TensorFlowCoursera · 2019

Let's talk

Open to interesting problems in agentic AI, retrieval, and edge vision. The fastest way to reach me is email.

ierolsen@outlook.comGitHubLinkedIn