-
🔭 I'm currently working on unreal engine and building autonomous AI agents on local hardware
-
🌱 I'm currently learning tensorflow, C++ and local LLM inference on edge hardware
-
👯 I'm looking to collaborate on autonomous AI systems, local LLM pipelines and anyone who likes to share knowledge
-
🤝 I'm looking for help with blueprint and Apple Silicon MLX inference optimization
-
💬 Ask me about unreal engine, ANF — Autonomous Native Forge, Node.js native modules, vLLM on Blackwell, python and everything about IT
-
📫 How to reach me turgaysavaci@gmail.com
A cloud-free, self-healing software production pipeline powered by 4 AI agents.
Zero middleware. Zero external dependencies. Pure Node.js native.
PM Agent → Architect Agent → Coder Agent → Reviewer Agent
↑__________________________________|
Self-Healing Loop
4 AI agents collaborate to analyze, design, write, and audit code —
running on local hardware with local LLM inference, completely offline.
Currently running on: NVIDIA Blackwell GB10 + DeepSeek-R1-32B
Porting to: Apple Silicon (Unified Memory) · ASUS Ascend (NPU)
// What I don't do
const express = require('express'); // ❌
const axios = require('axios'); // ❌
const _ = require('lodash'); // ❌
// What I do
const https = require('https'); // ✅
const fs = require('fs/promises'); // ✅
const { EventEmitter } = require('events'); // ✅Native Node.js only. If the runtime has it built-in, I don't install a package for it.
This isn't dogma — it's a deliberate choice to understand every layer of the stack.
I don't build in private and launch with a polished demo.
I build in public and document every failure along the way.
Every session produces:
- 🔴 What broke and why
- 🟡 What was attempted
- 🟢 What was learned
All of it lives in DEVLOG.md — updated after every session.