Tudor Iustin

AI Engineer · Generative AI · LLM Systems

01 / Profile

AI Engineer specialising in large language models, post-training, AI agents, and scalable applications powered by modern generative AI.

PyTorchFastAPI RAGAMD ROCm LangChainDocker SQL

Portfolio

Latest Projects

View all →

Open-source dataset · 01

Maple

A 16,000-sample full-stack code dataset for code generation, post-training, instruction tuning, research, and evaluation.

View dataset →

Open-source AI · 02

Donovan Agent

A terminal-native agentic assistant that can inspect code, edit files, run commands, browse the web, connect tools, and remember project context.

View project →

Track record

Experience

From native mobile products to production machine-learning and generative AI infrastructure.

May 2026 — Present

AMD

Generative AI Engineer · AMD AI Developer

Building generative AI applications, RAG pipelines, model-serving APIs, and hardware-aware PyTorch inference optimised for AMD CPUs and GPUs.

Nov 2025 — Feb 2026

Outlier

Artificial Intelligence Engineer

Contributed AI engineering expertise to the development and evaluation of intelligent systems.

May 2024 — Apr 2025

Glovo

Junior Machine Learning Engineer

Built real-time ML inference services with FastAPI and worked across preprocessing, feature engineering, model training, evaluation, and production monitoring.

May 2022 — Feb 2025

Independent · Upwork

iOS Developer

Delivered Swift and SwiftUI applications for 12+ clients across finance, retail, and e-commerce, achieving a client satisfaction rate above 95%.

Profile

About

AI engineering grounded in software development, product thinking, and production-ready systems.

What I do

Engineering useful intelligence.

AI Engineer specializing in Large Language Models (LLMs), post-training, PyTorch, AI agents, and scalable AI systems, building production-ready applications powered by modern generative AI.

Education

Computer Science & Data Science

BSc Computer Science from the University of Greenwich and the MicroMasters Program in Statistics and Data Science at MIT.

Languages

English, Romanian & Spanish

Native or bilingual English, native Romanian, and professional working proficiency in Spanish.

Connect