Julius Hietala
Machine Learning Software Engineer
Email: julius@juliushietala.com
Website: juliushietala.com
LinkedIn: linkedin.com/in/julius-hietala-8967b8a2
GitHub: github.com/hietalajulius
U.S. Permanent Resident (EB-1)
About
I'm a Machine Learning Software Engineer working on applied ML systems at the intersection of perception, control, and real-time inference. I focus on building ML that actually runs in production and integrates with software and control stacks.
I've shipped production ML systems, written technical tutorials on deep learning and computer vision, and built simulation environments for embodied AI. I've published peer-reviewed work at IROS (Best Paper Finalist) and in the International Journal of Robotics Research (IJRR) on robotic cloth manipulation, visual feedback control, sim-to-real pipelines, and benchmark development.
I'm currently interested in Vision-Language-Action (VLA) models with real-time controllers in the loop, targeting practical hardware rather than idealized execution.
Core Areas
- Python and PyTorch ML systems
- Computer vision and perception pipelines
- ML integration with simulation and control
- Research-to-production engineering
Experience
Nimble Robotics
Software Engineer
Working on deploying state-of-the-art ML and RL algorithms in production. Building intelligent robots that can pick and pack anything.
Smartly.io
Software Engineer
Led award-winning efforts to build solutions for automatic image and video generation and ad optimization. Helped some of the world's largest advertisers in scaling their social media advertising by integrating them into the Smartly.io advertising ML solutions.
Stray Robots
Co-Founder, CTO
Tools and services for 2D and 3D computer vision ranging from labeling to custom algorithms.
hanaa.io
Co-Founder, CTO
Easily buy and sell motorcycles through our service. Designed to remove barriers between sellers and buyers.
Junction
Co-Founder, Board Member
One of the largest hackathons in the world tackling real-world challenges with innovative tech solutions.
Publications
A Dataset and Benchmark for Robotic Cloth Unfolding Grasp Selection: The ICRA 2024 Cloth Competition - Published in the International Journal of Robotics Research (IJRR), January 2026.
Learning Visual Feedback Control for Dynamic Cloth Folding - Best Paper Award Finalist at 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2022), August 2022.
Technical Skills
Languages
Python, Rust, C++, TypeScript, JavaScript
ML Frameworks
PyTorch, Core ML, ExecuTorch
Computer Vision
Object Detection, Segmentation, Visual Control, 3D Vision
Systems & Tools
React, Next.js, ROS, Docker, Kubernetes
Research Areas
Reinforcement Learning, Sim-to-Real, VLA Models, Robotics