Mostafa Kermani Nia mostafa-kermaninia

👋 About Me

I'm Mostafa Kermani Nia, a Computer Engineering student at the University of Tehran.

My main interests lie at the intersection of machine learning and computer systems — particularly the infrastructure, data, and software systems required to build intelligent applications beyond isolated models and notebooks.

I enjoy working across the stack:

data pipelines → machine learning → LLM/agent systems → distributed infrastructure

with a particular interest in building systems that are scalable, reproducible, efficient, and useful in the real world.

🧭 Focus

                ┌────────────────────┐
                │        DATA        │
                │ ETL · SQL · Search │
                └─────────┬──────────┘
                          │
                          ▼
                ┌────────────────────┐
                │  MACHINE LEARNING  │
                │ Models · Features  │
                └─────────┬──────────┘
                          │
                          ▼
                ┌────────────────────┐
                │     AI SYSTEMS     │
                │ LLMs · Agents · ML │
                └─────────┬──────────┘
                          │
                          ▼
                ┌────────────────────┐
                │   INFRASTRUCTURE   │
                │ Distributed · MLOps│
                └────────────────────┘

🚀 Selected Work

🤖 Orbit — Local AI Agent

A local-first AI system combining LLMs, speech, persistent memory, tool calling, automation, and real-time interaction.

LLM Systems Agents Local AI Python

📊 Math Misconception Analytics Pipeline

An end-to-end data and ML pipeline covering ETL, relational storage, feature engineering, containerization, and reproducible workflows.

Data Engineering ML Docker CI/CD

🎙️ Speech Processing Toolkit

An applied machine-learning system for audio representation, feature engineering, classification, and speaker-related modeling.

Applied ML Audio AI TensorFlow scikit-learn

🌐 P4 In-Band Telemetry Analyzer

A programmable telemetry system for real-time network measurement, congestion analysis, and stateful data-plane monitoring.

Systems Telemetry P4 Networking

↓

More projects in ML, data processing, parallel computing, AI systems, and systems engineering are available across my repositories.

⚙️ Engineering Stack

Languages

ML & Data

Infrastructure & Systems

ML / AI	Data	Systems	AI Engineering
PyTorch TensorFlow scikit-learn XGBoost	Pandas NumPy SQL SQLAlchemy	Docker Kubernetes Linux Networking	LLMs Tool Calling Agents Local Inference

🔬 Areas I'm Exploring

research_interests = {
    "machine_learning": [
        "representation learning",
        "applied ML",
        "ML systems",
    ],
    "data_systems": [
        "data-intensive applications",
        "data pipelines",
        "large-scale processing",
    ],
    "ai_systems": [
        "LLM systems",
        "AI agents",
        "local & efficient inference",
    ],
    "distributed_systems": [
        "distributed computing",
        "parallel processing",
        "systems for ML",
    ],
}

📈 GitHub

🎓 Beyond the Code

I study Computer Engineering at the University of Tehran and have been involved in teaching and academic activities alongside my engineering work.

For academic background, teaching experience, awards, and other details:

📄 View my CV

🤝 Connect

I'm always interested in conversations around Machine Learning · Data Systems · AI Infrastructure · Distributed Systems

_{Data → Intelligence → Systems

Building at the intersection of machine learning and scalable software.}

Navigation Menu