Enes Gül

Journal

Log

Weekly notes on the work, tools, and events I have been spending time with.

Week 40

  1. Built my portfolio

    Built a new personal portfolio with strong visual inspiration from Onur’s site, while creating my own structure around projects, journey, writing, notes, logs, stack, and about.

  2. Published the portfolio repository

    Cleaned the repository history and published the portfolio as a public GitHub project.

    Repo

  3. Built the Notes archive

    Added PDF-backed collections for Distributed Systems, Cloud Architecture, Operating Systems, and Computer Networks.

  4. Added my existing writing

    Added my two Medium articles to the Writing archive alongside the technical notes.

  5. Continued Far Away from Codex

    Kept developing the product direction and architecture while the broader agent-attention research continued.

Week 39

  1. AWS Cloud & AI Day Istanbul — Sep 23

    Attended AWS Cloud & AI Day and joined sessions around AI-assisted software development, including “Introducing AI-Driven Development Lifecycle (AI-DLC).”

    AWS Cloud & AI Day in Istanbul
    AWS Cloud & AI Day · Istanbul, 2026

  2. Oracle OpenGrok contribution

    Finished a small contribution that changes Bazaar repository tests to exercise real Bazaar/bzr behavior instead of relying only on handcrafted parser input. The PR is open and awaiting maintainer workflow approval.

    PR

  3. Far Away from Codex research

    Continued working through agent integration, attention flows, permission round-trips, reconnect behavior, and Telegram as the first remote interaction channel.

Week 38

  1. Local LLM experiments

    Set up Ollama and connected local models to VS Code through Continue. Tested coding workflows and hit the practical limits of long context and CPU-only inference.

  2. Started expanding Far Away from Codex

    Moved the project beyond simple completion notifications toward an attention layer for coding agents: approvals, input requests, failures, and bounded remote responses.

    Repo

  3. Read through ACP

    Studied Agent Communication Protocol and compared where a standard protocol makes sense versus native agent hooks.

    Source

  4. Kaggriculture experiment

    Tried an agent/model approach for the Kaggriculture competition. The initial results weren’t promising enough, so I stopped working on it.

Week 34

  1. Roketsan LLM/RAG Workshop

    Built a small RAG workflow with Gemini embeddings, FAISS retrieval, and LangChain while experimenting with agent tools and retrieval quality.

    Repo

  2. Roketsan Computer Vision Workshop

    Worked through image processing, classification, object detection, and tracking exercises using OpenCV, MobileNetV2, YOLO11, and ByteTrack.

    Repo