Engineer | Research | GenAI | MLOps | Embedded Systems
I am an engineer focused on building complete systems.
I work across hardware, firmware, software, AI/ML, and infrastructure, and take ideas from concept to production with an emphasis on correctness, scalability, and long-term maintainability.
| Project Name | Description | Link |
|---|---|---|
| ErrorHunter | A popular, beginner-friendly bugfix challenge repository that helps developers learn Python debugging and open-source contribution through test-driven development. | Repo |
| PDF_Spool_Optimizer | A Python CLI utility that rasterizes and compresses complex PDFs to eliminate print spooling bottlenecks. | Repo |
| BERT_QA_Engine | Trained Transformer (Fine Tuned) Model | Repo |
| Advanced Object Tracking & Detection in Video Streans | A PyTorch implementation of an advanced pedestrian tracking pipeline on the MOT20 dataset. Features a fine-tuned Faster R-CNN (ResNet-50-FPN) detector coupled with an Adaptive Kalman Filter and Temporal Consistency checks for robust tracking in crowded scenes. | Repo |
| Named Entity Recognition | A context-aware Named Entity Recognition (NER) system for detecting and sanitizing Personally Identifiable Information (PII) in unstructured text logs using fine-tuned DistilBERT. | Repo |
| Java-Core-to-Advanced | A structured learning repository that goes beyond standard tutorials to teach the internal mechanics and architectural thinking behind Java code. | Repo |
| Chip_Design_Roadmap | A structured repository reflecting your lowest-level systems engineering foundation, focusing on hardware and embedded systems roadmaps. | Repo |
| VLSI Opensource Tools Install Master Script | A Single Repo contains the VLSI flow (1.Semi Custom , 2. Full Custom) and script to install various open-source tools for VLSI design, including Yosys, OpenROAD, Magic, KLayout, and NGspice. | Repo |
My foundation started at the lowest levels of systems: registers, signals, timing, and constraints.
Over time, that understanding expanded upward:
- from firmware to embedded systems
- from systems to software
- from software to machine learning
- and into production-grade AI systems
Today, my focus is on building systems that not only work,
but continue to operate reliably as complexity and scale increase.
- Machine Learning and AI model development
- Production ML using MLOps practices
- Backend system design and implementation
- Embedded systems and firmware development
- Technical leadership and team mentorship
I prioritize engineering fundamentals, clear abstractions, and system-level thinking.
My current work is centered on:
- Generative AI systems
- Production-grade backend architectures
- End-to-end ML systems
(training → deployment → monitoring → iteration)
Models are experiments.
Systems are responsibilities.
I treat machine learning models as long-lived products, not notebooks.
- Experiment tracking and model lifecycle management
- Data, model, and code versioning
- CI/CD for ML workflows
- API-based model serving
- Monitoring, drift detection, and observability
The objective is consistent:
predictable behavior in unpredictable environments.
My embedded work spans:
- Arduino, ESP32, Raspberry Pi
- STM32 microcontrollers
- Register-level programming
- RTOS and Embedded Linux
I am particularly interested in systems where hardware, software, and intelligence intersect.
Engineering impact extends beyond code.
- Lead GenAI and Embedded Systems teams
- Mentor and train engineering students and professionals
- Active in the engineering community in Coimbatore, Tamil Nadu
- Develop structured learning roadmaps for hardware and software engineers
I value building environments where engineers grow in both skill and judgment.
- Maintain repositories with engineering roadmaps and reference implementations
- Contribute to and support open-source projects
- Emphasize clarity, reproducibility, and real-world applicability
If something helped me understand a system deeply,
I document it so others can build on it.
- End-to-end ownership
- Strong fundamentals
- Thoughtful mentorship
- Systems that work beyond demos
Good engineering is quiet.
It fails less, surprises less, and lasts longer.
📫 Open to collaboration and engineering discussions.
Mathi Yuvarajan
Engineer
✉️ mathiyuvarajan.dev@gmail.com | https://techwithmathi.in