Mid-level full-stack and IoT developer. Currently a tester and project manager at Debtrak, working remotely with an Australian team. MSc in Computer Science with a published numerical method for solving integral equations using support vector regression.
I am currently working as a tester and project manager at Debtrak, remotely. I hold a Master's in Computer Science with a publication in Computational and Applied Mathematics, where we presented a new numerical method for solving integral equations using support vector machines.
Day to day I am a mid-level full-stack developer, participating as a developer in a software development team across a range of projects. Alongside that I design and build IoT devices as a freelancer and hobbyist, taking them from schematic to working hardware.
Milad Hasani
Full-stack Developer · IoT Device Developer
Product ownership on a legacy-to-web migration, freelance hardware work, and full-stack delivery inside a software team.
01/2022 — Present
Australia (remote)
01/2022 — Present
Debtrak— Debt collection software vendor
Australia (remote)
04/2022 — 01/2023
Tehran, Iran
04/2022 — 01/2023
Self-employed— Freelance
Tehran, Iran
Extended my knowledge and experience in IoT engineering by carrying out related small-scale projects from scratch.
01/2018 — 2022
Tehran, Iran
01/2018 — 2022
Fusoft— Software development company
Tehran, Iran
Pulled live from GitHub, newest push first, alongside hardware that never lived in a repository.
Designed and manufactured a smart greenhouse controller from scratch: a device combining soil and air humidity, temperature and light sensors with the matching actuators, to control a plant habitat intelligently and remotely.
10/2016 - 12/2016
Android clients plus protocol notes. The tracker server is a separate app from wg-quota; both happen to run in one MikroTik container on different ports (9090 vs 8080).
A Gradio app that takes a person photo and a garment photo, then generates a try-on image with pretrained CatVTON (ICLR 2025). Built for a laptop RTX 4070 (~8 GB VRAM).
A Kotlin project. See the repository for details.
A Kotlin project. See the repository for details.
A Python project. See the repository for details.
A Python project. See the repository for details.
A JavaScript project. See the repository for details.
A Jupyter Notebook project. See the repository for details.
My solutions to TensorTonic problems
A JavaScript project. See the repository for details.
Developing multi layer perceptron neural network from scratch
A stack that spans a .NET and Angular web career, Python for numerical and machine learning work, and the firmware side of connected devices.
My MSc thesis turned into a peer-reviewed method for solving integral equations with support vector regression.
K. Parand, M. Hasani, M. Jani, H. Yari
Computational and Applied Mathematics, Springer · Volume 40, Issue 7
In this paper, a new method based on least squares support vector regression (LS-SVR) is presented as a numerical method for solving linear and nonlinear Volterra-Fredholm integral equations.
DOI 10.1007/s40314-021-01471-0
Scientific computing at Master's level, software engineering at Bachelor's level, and a robotics contest that started it all.
Scientific Computing
Shahid Beheshti University
Shahrood University of Technology
Placed 3rd among 21 teams in the line follower / medic robot major during my post-secondary education.