close
Skip to content
View faliye's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report faliye

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
faliye/README.md

Hi there, I'm faliye 👋

"コーディングできない、AI Agent最高"

I am a system-level full-stack engineer and researcher, currently based in Japan. My work spans from the lowest hardware/OS kernel layers to AI infrastructure and modern web security architectures.

🇺🇸 English (Default)
  • Linux Kernel & Hardware Tuning: Active contributor to the Linux Kernel. Debugging complex concurrency issues (Lost Wakeup, D-state deadlocks), tuning PCIe/Thunderbolt 5 hardware protocols, and writing patches for i2c, btrfs, and vga. (18+ patches merged into mainline/stable).
  • AI Infrastructure & Local Compute: Squeezing hardware limits for LLMs. Optimizing cross-node RPC bandwidth, vLLM distributed inference, and building high-performance local AI clusters (RTX 5090/5080).
  • Full-Stack & Security Architecture: Building Bot defense systems, anti-scraping architectures, and modern web applications using Micro-frontends (React) and AI Agents (RAG / MCP).
🇯🇵 日本語

日本を拠点に活動するシステム級フルスタックエンジニア・リサーチャーです。ハードウェア/OSカーネルの最下層から、AIインフラ、モダンWebアーキテクチャまで一気通貫で手掛けています。

  • Linuxカーネル & ハードウェアチューニング: 複雑な並行処理バグ(Lost Wakeup、D状態のデッドロック等)のデバッグ、Thunderbolt 5プロトコルの極限チューニング。i2c、btrfs、vga等のサブシステムへ公式パッチを提供 (18件以上がアップストリームにマージ済)。
  • AIインフラ & ローカル算力基盤: LLM推論におけるハードウェア限界の追求。vLLM分散推論、クロスノードRPC帯域の最適化、マルチGPUクラスタ(RTX 5090/5080)の構築。
  • フルスタック & セキュリティアーキテクチャ: ボット防御・スクレイピング対策アーキテクチャの構築、マイクロフロントエンド(React)やAI Agent (RAG/MCP) のプロダクト実装。
🇨🇳 中文

Base 在日本的系统级全栈工程师与研究员。技术栈贯穿从最底层的硬件/OS内核,到AI算力基础设施,再到现代 Web 与安全架构。

  • Linux 内核与硬件调优: 解决复杂的并发死锁(Lost Wakeup, D-state),Thunderbolt 5 协议极限调优。为 i2c、btrfs、vga 等子系统编写补丁 (向 Linux 上游主线贡献并合入 18+ Patches)。
  • AI 基建与本地算力: 压榨大模型推理的硬件极限。vLLM 分布式推理优化,跨节点 RPC 通信调优,以及自建多卡算力集群(RTX 5090/5080)。
  • 全栈与业务安全架构: 高并发防爬虫与 BOT 防御体系,微前端 (React) 架构,以及 AI Agent (RAG/MCP) 的工程化落地。

📝 Technical Writings & Proof of Work

I regularly document my hardcore debugging logs, hardware benchmarks, and architectural designs (Mostly in Japanese):


“BUG BUG BUG BUG”

Pinned Loading

  1. novel-translate-tts novel-translate-tts Public

    Python

  2. thunderbolt-nhi-re-kick-TX-ring-doorbell-on-missed-p thunderbolt-nhi-re-kick-TX-ring-doorbell-on-missed-p Public

    Shell

  3. spd-fix-lost-wakeup-when-waiting-for-a-muxed-region-p spd-fix-lost-wakeup-when-waiting-for-a-muxed-region-p Public

  4. nwbutton-fix-lost-wakeup-in-button_read-p nwbutton-fix-lost-wakeup-in-button_read-p Public

  5. singlefs singlefs Public

    Rust

  6. singlefs-ai-sop-en singlefs-ai-sop-en Public

    Shell