About · AI Engineer & Developer
AI developer and AI engineer focused on local AI, local LLMs, RAG systems, and AI agents — building evaluated, verifiable systems that keep humans in control of the decisions that matter.
One Name, Many Spellings
Hussain Nazary is the canonical spelling I use. The following are common English transliterations of the same Persian/Dari name (حسین نظری) and all refer to me:
The first name, حسین (Husayn), is romanized several ways because Arabic and Persian vowel sounds have no single English equivalent:
The last name, نظری, is also romanized more than one way depending on dialect and transliteration convention:
Because transliteration is not standardized, the five given-name spellings combine with the three family-name spellings in everyday writing. Whether you search for Hussein Nazary, Hussain Nazari, or Hussein Nizari, you are looking for the same engineer — this site and its structured data make that equivalence explicit for search engines, AI assistants, and knowledge graphs.
Who I Am
I'm Hussain Nazary, an AI engineer focused on one question: which parts of your team's knowledge work should software be doing instead?
My specialty is document- and information-intensive workflows — research, review, retrieval, reporting — where the steps repeat and the answers have to be verifiable. Since 2024 I've designed and shipped complete systems for exactly this kind of work: evaluating models honestly, building for production constraints, and treating "works in a demo" as nowhere near good enough.
As a Python developer and AI engineer, I work across the full stack of modern AI — local LLMs running through llama.cpp and Ollama, RAG systems with hybrid search and vector databases, AI agents orchestrated with LangGraph, and the on-premise deployment layer that keeps sensitive data inside your infrastructure. My open-source project GGUF Loader (GGUFLoader) is that deployment layer, made tangible.
Before AI, I led technical operations in peer-to-peer trading (2018–2021), where software had to survive real money and real pressure. I hold every automation to that standard: evaluated trade-offs, deliberate architecture, and systems that keep working after launch.
Areas of Work
The technology follows the workflow — but these are the areas I live in day to day.
Running language models on-device with GGUF, llama.cpp, and Ollama — when data must never leave your walls.
Retrieval-augmented generation with hybrid search, vector databases, and cited, verifiable answers.
Multi-step, tool-using agents orchestrated with LangGraph — drafted, evaluated, and escalated to humans.
Production Python for AI systems — from raw-PyTorch training loops to deployable applications.
Creator of GGUFLoader, the zero-CLI desktop engine for running GGUF models locally.
Private AI infrastructure for regulated industries — architecture where privacy is built in, not bolted on.
Projects
Private AI infrastructure: a desktop engine that runs GGUF-format language models entirely on local hardware — drag-and-drop loading, live RAM/VRAM dashboard, plugin architecture.
Legal research automation: a fully on-device RAG pipeline (BGE-M3 embeddings, ChromaDB, BM25 hybrid search, Ollama) that answers questions from a firm's own documents with verified citations.
Everyday workflows automated with on-device AI: a voice-first Android assistant replacing calendar, reminders, expense tracking, and location alerts in one app.
Smaller tools and experiments — LLM-Toolkit, raw-pytorch-minigpt, Mobile AI Assistant — live on GitHub →
Common Questions
Hussain Nazary is an AI developer and AI engineer who automates document-heavy and knowledge-intensive workflows with evaluated, verifiable AI systems. His work centers on local AI, local LLMs, RAG systems, AI agents, and on-premise deployment. He is the creator of GGUF Loader (GGUFLoader), Lawyer Assistant, and GPT Calendar.
Yes. Hussein Nazary and Hussain Nazary are the same person. "Hussain" is the canonical spelling; "Hussein" is a common alternate romanization of the same Persian/Dari given name, حسین.
Yes. Hussain Nazari and Hussain Nazary are the same person. "Nazary" and "Nazari" are two common romanizations of the same Persian/Dari family name, نظری.
Yes. Hossein Nazary is the same person as Hussain Nazary. "Hossein" and "Hussain" are both English transliterations of the Arabic/Persian given name حسین (Husayn).
Hussain Nazary created GGUF Loader (GGUFLoader), a zero-CLI desktop engine for running language models locally; Lawyer Assistant, a privacy-first legal research assistant built on a local RAG pipeline; GPT Calendar, a voice-first on-device AI assistant; and raw-pytorch-minigpt, a GPT-style LLM built from scratch in raw PyTorch.
GGUF Loader (written GGUFLoader) is an open-source desktop engine created by Hussain Nazary that runs GGUF-format language models entirely on local hardware — no terminal, no cloud. It features drag-and-drop model loading, a live RAM/VRAM dashboard, and a plugin architecture. Read the case study.
The name حسین نظری is written in Arabic script and must be converted into the Latin alphabet for English. That conversion — transliteration — has no single fixed rule, so the same name commonly appears as Hussain/Hussein/Husain/Hossein for the given name and Nazary/Nazari/Nizari for the family name. Every combination refers to the same person: Hussain Nazary.
Keep Exploring
Services, process, and selected work in one place.
Deep dives on local AI, RAG, agents, and private deployment.
The file format behind GGUF Loader and local AI.
How this site helps AI assistants understand who I am.
Tell me what your team repeatedly does and where the bottleneck is — you'll get an honest assessment of whether AI is a good fit.
Describe Your Workflow