Email: md.ziaulhasan00@gmail.com
Phone: +880 1741144605

AI Engineer with a strong grasp of modern AI concepts including LLM Finetuning (SFT, LoRA, QLoRA, RLHF, DPO) and AI Agent Development. Have a solid understanding of major neural network architectures (Transformers, Diffusion models, CNNs, GANs, RNNs, LSTMs etc.), can implement them in PyTorch and can customize to meet problem needs. Proficient in Python and its ecosystem including PyTorch, LangChain, LlamaIndex, NumPy, Pandas, Matplotlib, Scikit-learn, FastAPI. Passionate about implementing cutting-edge AI research from scratch. My technical focus is centered on task-specific LLM fine-tuning, semi-autonomous agentic systems, and optimizing model alignment through Human-in-the-Loop data generation.

Advanced AI Trainer
Invisible Technologies | July 2025 - Present
At Invisible I am currently focused on LLM finetuning, where I manage a portfolio of high-impact projects. These include a primary
initiative in MCP-based tool calling, Reinforcement Learning from Human Feedback (RLHF), and the specialized training of reasoning models.

AI Research Engineer
AIMS Lab, United International University | January 2025 - June 2025
My focus of research was the application of AI in Healthcare. I have worked on building methodologies to detect neurodevelopment disorder
of children from parent-AI conversation.

AI Data Engineer
Outlier | February 2024 - January 2025
I was an Independent Contractor at Outlier AI. Contributed to the development of datasets for training large code models, with a focus on JavaScript and Python.

Competition Expert
Kaggle | May 2023 - January 2024
Earned two bronze medals in "CAFA 5 Protein Function Prediction" and "NeurIPS 2024 - Predict New Medicines with BELKA," attaining
Competition Expert status.Showcased my ability in tackling complex real-world problems.

Business Analyst Intern
Tork | February 2023 - April 2023
Supported the branding and marketing team with data-driven insights. Applied analytical skills to guide strategic decisions in a real-world business environment.

BNID – Balanced Network for Imbalanced Data
Designed a new neural network architecture that tackles data imbalance not through augmentation, but by making the architecture itself inherently robust to imbalance.

Aitist – AI Model for SVG-Based Image Generation
Building a language-to-code model that generates SVG images from natural language instructions. The goal is to power a new kind of digital canvas where designers describe ideas in words, and AI handles the visuals.

Protein Function Prediction from Amino Acid Sequences
Developed a deep learning-based model to predict biological functions of proteins from raw amino acid sequences using neural networks.

Bachelor of Science in Biomedical Engineering
CUET | February 2020 - January 2025
CGPA: 3.76/4.00

Micro BSc in Computer Science
Phitron | February 2022 - February 2023
CGPA: 4.00/4.00

Languages: Python, JavaScript
Data Analysis: Numpy, Pandas, Matplotlib, SciPy
ML Frameworks: PyTorch, Scikit-Learn
GenAI Frameworks: LangChain, LangGraph, LlamaIndex
Frontend: React
Backend: Django