yiannisha GithubSTATS3 FAILED STARTUPS0 SUCCESS STORIESCOUNTLESS MORE FAILURESCURRENT OBSESSIONSML&RL, CUDA, RUST, GPU Architecture,
real-world operations research problems for warehouses and factories,
continual learning, memory systems and robotic agentsTHINGS IM GOOD AT-
ML, RUST, ROBOTICS, CUDA
I'm a software engineer mostly focused on ML systems, robotics software, inference optimization, and low-level performance work.
Right now I work at Tools for Humanity as an ML and Rust engineer, working on robotics, optimizing GPU kernels, and building ML pipelines for Jetsons.
I also do independent research around memory, continual learning, and real-world OR problems. You can find some of my research work on my personal research feed and Synkrasis Labs.
I AM OBSESSED WITH WAREHOUSES AND FACTORIES
I am obsessed with writing software, solving hard problems, building systems
and I will probably keep doing it until I die.I FOUND MY CRAFT AND MY CRAFT HAS FOUND ME
MY ONLY GOAL IN LIFE SHALL BE TO MASTER IT
Work Experience
A Bit Of My Story
Started programming in high school, mostly through gamedev and minecraft mods.
Did ~14 months of military service as an EOD, learned python in parallel, and got my first freelance programming work in the meantime.
Then I got into the ECE School of NTUA doing a MEng.
After around 1.5 years of studying I dropped out to work full-time in startups.
I ended up working across energy, robotics, manufacturing, computer vision, AI systems, and infrastructure, while also co-founding companies and leading technical work.
Eventually I realized the real constant was not startups or products, it was the need to learn hard things and solve problems at a high level.
That's basically what I optimize for now.
ML, robotics, systems, CUDA, research, and problems that force me to get better.
Also still trying to finish NTUA at some point.
PhD is not out of the question either.
Some projects I've worked on:
Get in touch:
me@yiannisha.dev
I'm a software engineer mostly focused on ML systems, robotics software, inference optimization, and low-level performance work.
Right now I work at Tools for Humanity as an ML and Rust engineer, working on robotics, optimizing GPU kernels, and building ML pipelines for Jetsons.
I also do independent research around memory, continual learning, and real-world OR problems. You can find some of my research work on my personal research feed and Synkrasis Labs.
I AM OBSESSED WITH WAREHOUSES AND FACTORIES
I am obsessed with writing software, solving hard problems, building systemsand I will probably keep doing it until I die.I FOUND MY CRAFT AND MY CRAFT HAS FOUND ME
MY ONLY GOAL IN LIFE SHALL BE TO MASTER IT
Work Experience
Tools for Humanity - ML & Rust Engineer Mar 2026 - Present
Working on robotics systems, ML infrastructure, and performance-critical software. Optimizing GPU kernels and ML pipelines for Jetson deployments. Building Rust and ML systems around real-world robotics constraints. Rust, CUDA/C++, PyTorch, Jetson, Robotics
Progressive Robotics - Machine Learning & Robotics Engineer Oct 2025 - Mar 2026
Optimized CPU-bound RL environments, decreasing training time by around 2x. Implemented GPU-optimized physics heuristic algorithms in CUDA to enable GPU-native training. Designed the pretraining stack, synthetic data generation pipelines, and transformer-based physics-aware models. PyTorch, CUDA/C++, Airflow, Zarr, GCP
Delian Alliance Industries - Backend/Infra & Robotics Software Engineer Feb 2025 - Oct 2025
Designed and implemented the teleoperation streaming stack from scratch, lowering latency to ~150ms while making the stream much smoother. Built custom monitoring / service management infrastructure for deployed devices. Maintained the core gRPC library used across Olympus services and optimized detection models for NVIDIA Jetson deployments. C++, Python, gRPC, ROS2, TensorRT, WebRTC, mediasoup, gstreamer
NAS-AI - Co-Founder & CTO Feb 2024 - Jun 2024
Designed and deployed a real-time video streaming and object recognition product for manufacturing lines. Improved client efficiency and reduced waste using edge-deployed CV algorithms. Built a lightweight robotics stack and a custom OTA update service for deployed devices. Python (OpenCV), Rust, MQTT, ZQM, C++, Docker, Raspberry Pi
IQ3Solar - Full-Stack & Data Engineer Jul 2023 - Jan 2024
Automated collection, processing, and storage of PV data. Deployed ML algorithms with automated retraining and finetuning for individual systems. Built a full-stack platform for tracking, monitoring, and analyzing PV-related data. Python, XGBoost, AWS
A Bit Of My Story
Started programming in high school, mostly through gamedev and minecraft mods.
Did ~14 months of military service as an EOD, learned python in parallel, and got my first freelance programming work in the meantime.
Then I got into the ECE School of NTUA doing a MEng.
After around 1.5 years of studying I dropped out to work full-time in startups.
I ended up working across energy, robotics, manufacturing, computer vision, AI systems, and infrastructure, while also co-founding companies and leading technical work.
Eventually I realized the real constant was not startups or products, it was the need to learn hard things and solve problems at a high level.
That's basically what I optimize for now.
ML, robotics, systems, CUDA, research, and problems that force me to get better.
Also still trying to finish NTUA at some point.
PhD is not out of the question either.
Some projects I've worked on:
llm-tool
A Python module that automatically turns functions into definitions for LLM tool calling. Built with Rust for fast string parsing and low overhead. Rust, Python
zspotify - Top Contributor
Spotify song downloader without injecting into the Windows client. Reached #1 trending on GitHub for a week. Development stopped after a cease and desist letter from Spotify. Python, Ffmpeg
RobinStore - Best MVP Award Mar 2023
Automated supermarket product scraping tool with real-time accurate data from all over Greece. Proprietary product matching based on image and text embedding similarity. Won Best MVP at the VAIM 2023 startup competition. Python, JS, Embeddings
Get in touch:
me@yiannisha.dev