I am a Mechatronics Engineering Student passionate in software development and robotics, along with creating innovative digital experiences with modern technologies and AI.
About Me
I'm a Mechatronics Engineering student at the University of Waterloo with a strong foundation in engineering principles and hands-on project design. My passion lies in software engineering, where I focus on creating automation systems, AI-driven applications, and scalable web solutions that bring real-world impact.
My Journey
Through leadership roles and extracurricular activities, I’ve learned how to work across teams and lead projects from concept to completion. My experience spans full-stack development, AI/robotics, and mechanical design — giving me a unique perspective at the intersection of hardware and software.
I’m motivated by the challenge of solving complex problems through thoughtful design and clean, efficient code. My long-term goal is to grow as a software engineer, working on impactful projects in automation, intelligent systems, and innovative digital platforms.
Outside academics, I enjoy staying active through weight training and basketball, diving into books that expand my perspective, and traveling with my family to experience new cultures. These interests keep me balanced, creative, and constantly inspired.
Skills
A blend of frontend, backend, hardware, and AI expertise developed through projects, internships, and hands-on engineering work.
Languages & Frameworks
Python, TypeScript, JavaScript, SQL, Java, C/C++, HTML/CSS, Bash. React, Next.js, Node.js, FastAPI, Flask, TensorFlow, PyTorch, Pandas, NumPy, Tailwind CSS, Framer Motion, OpenCV, MediaPipe, OCR.
Cloud Infrastructure & DevOps
Google Cloud Platform (Vertex AI, BigQuery, Cloud Run), AWS, Docker, PostgreSQL, Supabase, Git, Linux, Elasticsearch, MongoDB, WebSockets, Cloudflare.
AI & Data Tools
OpenAI API, LangChain, Google Gemini, Pgvector, Vertex AI SDKs, Claude Code, Cursor, Vector Embeddings, Fivetran, REST APIs, ROS 2.
Mechanical & Hardware Engineering
AutoCAD, SolidWorks, PCB design, circuit design, and machine shop tools for hardware integration.
Professional Experience
My journey through various roles and teams, building impactful digital experiences
Machine Learning Research & Development Intern
Perceivable Design Studios Inc.
Leveraging extensive machine learning experience to research and develop multimodal translation frameworks that bridge ASL and speech, with a focus on real-time performance optimization and model deployment.
Key Achievements:
- Researched and developed a multimodal translation framework in Python to bridge ASL and speech; engineered a streaming ASR pipeline by fine-tuning Whisper via Torchaudio, optimizing the model for low-resource environments through iterative hyperparameter tuning and NumPy-based signal processing
- Engineered custom PyTorch Transformer architectures to model complex temporal dependencies in sign language, utilizing MediaPipe skeletal hand-tracking data to achieve 95% gesture recognition accuracy across diverse datasets via advanced data augmentation and OpenCV preprocessing
- Optimized model inference and deployment cycles by evaluating the trade-offs between accuracy and latency; implemented ONNX quantization and WebSocket synchronization to reduce end-to-end lag by 40%, validating research benchmarks on edge hardware to ensure real-time performance
Software Engineering Intern
Friedmann AI
Currently developing and testing AI-powered chatbot for customer support along with other product features.
Key Achievements:
- Implemented complete Calendly integration and booking system with pre-filled forms, improving demo scheduling efficiency for 30+ potential clients
- Enhanced financial plan generator to meet FP Canada compliance requirements, developing a 1,200+ line prompt engineering system
- Built comprehensive advisor profile management system with API endpoints and settings UI, streamlining information management for financial advisors
- Created developer testing infrastructure including test signup endpoint and authentication bypass, reducing development workflow time by 75%
- Resolved critical UI/UX issues including CSP configurations, Stripe client initialization, and preload optimizations
Software Engineer Intern
Normative
Contributed to full-stack and AI development projects, enhancing internal tools and building AI-powered document processing systems.
Key Achievements:
- Built and enhanced Freedman's internal site using Retool and JavaScript, improving usability for 30+ team members
- Created an OCR-based document processing system with Python backend and HTML/CSS frontend, improving text recognition accuracy of scanned PDFs by 75%
- Developed and tested AI-driven business proposal generation using reinforcement learning and Python, optimizing proposal quality and reducing editing time
Autonomy Team Member
Waterloo Aerial Robotics Group (WARG)
Working on the autonomy team to advance drone navigation and perception systems using Python and machine learning.
Key Achievements:
- Developing computer vision pipelines with Python and ML models to increase drone navigation accuracy
- Optimizing autonomy algorithms for object detection, landing precision, and flight stability
- Collaborating with a multidisciplinary team to integrate autonomy modules into larger UAV systems
Featured Projects
Here are some of my recent projects that showcase my skills and passion for development.
Proof - Real-Time Polymarket Intelligence Extension
Chrome extension that adds real-time intelligence directly on Polymarket market pages. Won 2nd place in the Polymarket track at NexHacks 2026 @ CMU. Turns raw markets into informed decisions by surfacing correlation, context, and momentum where trades happen. Features related markets, trending markets, and real-world news integration.
Brain Battle - Interactive Quizzing and Study Platform
Production-ready platform for competitive brain training with real-time multiplayer capabilities. Features WebSocket-based live gameplay, dynamic scoring algorithms, and diverse question categories. Built with scalable microservices architecture, persistent leaderboards, and responsive design for engaging user experience.
Dermalens - AI-Powered Skincare Analysis
Production-ready platform leveraging Gemini API for medical-grade skin analysis, Elasticsearch for intelligent recommendations, and real-time data pipelines. Features multi-angle facial scanning, personalized routines, and scalable microservices architecture on Google Cloud Platform.
Coursely - Waterloo Elective Advisor
AI-powered platform helping engineering students choose electives through personalized recommendations. Built with Next.js, GPT-4o-mini, and RAG architecture featuring vector semantic search across 284+ courses. Includes natural language chat interface and intelligent course matching based on student profiles and career goals.
Mid Chats - AI-Powered Chat Application
Fully functioning personal chat application with real-time messaging capabilities. Features personal info storage for resume and cover letter building, along with comprehensive "study mode" function, intelligent conversation handling, message history, and seamless user experience. Built with modern full-stack technologies for scalable and responsive communication.
Global Solar Panel Energy Calculator
Web app that estimates annual solar output using NASA irradiance data. Python backend with a lightweight HTML/CSS front end; calculates per-panel output by location and panel efficiency.
Gym Posture Corrector (Real-Time Form Feedback)
Real-time posture analysis using OpenCV/MediaPipe to extract skeletal keypoints and a TensorFlow classifier to distinguish good vs. bad exercise form. Flask + JS UI for instant feedback.
OCR Image Scanner / Medical Document Analyzer
OCR pipeline with Python + OpenAI APIs to extract structured info from scanned PDFs; front end in HTML/CSS for quick review and export.
Nutrition/Fitness Plan Builder
Generates tailored workout + meal plans based on user goals and restrictions. Gemini API prompts via Flask backend with a simple web UI.
Tic Tac Toe Playing VEX Robot
A VEX robot capable of playing Tic Tac Toe against human opponents. Used computer vision to recognize game board states and mechanical systems for piece placement. Demonstrated integration of robotics, computer vision, and game theory algorithms.
Get In Touch
I'm always open to discussing new opportunities and interesting projects. Let's connect and create something amazing together.
Let's Connect
Whether you have a project in mind, want to collaborate, or just want to say hello, I'd love to hear from you.
stjeyara@uwaterloo.ca
Location
Available for remote work and willing to relocate
Availability
Open to new opportunities