About Me

About Echo Yin

I'm a senior AI engineer with over 8 years of experience building intelligent systems and machine learning solutions that transform how businesses operate. My journey in AI began with fascination for neural networks and has evolved into expertise in designing scalable, production-ready AI systems. I specialize in deploying machine learning models, with deep expertise in applied AI techniques like RAG, MCP and toolings. When I'm not tackling models, you can find me researching cutting-edge AI techniques, contributing to open-source ML projects, or mentoring aspiring engineers.

0 Years Experience
0 Projects Completed
0 Happy Clients

Experience

Nov. 2025 – Present

Senior A.I. Developer

Impact Assessment Agency of Canada / Agence d'évaluation d'impact du Canada

Impact Assessment Agency of Canada conducts federal impact assessments of major projects, like mines, ports, and dams, in federal jurisdiction

  • Collect and document clients' requirements and develop digital and infrastructure specifi cations for our A.I. enabled systems.
  • Design and coordinate the development, installation, integration and operation of our experimental A.I. systems.
  • Upgrade and develop Indigenous Consultation procedures for our Assessment Management System (AMS)
Jul. 2023 – Apr. 2025

Online Programmer

Brass Lion Entertainment

Brass Lion Entertainment is a game development studio dedicated to creating original fictional universes, characters, stories, and themes. Contributed to the development of online systems and live-ops features to enhance game operations.

  • Developed an item migration tool, optimizing item management by 96%
  • Created a server log download tool, reducing QA bug reporting time by 95%
  • Implemented a matchmaking system and data timer for server-side requests
  • Supported the Character Creation group with essential online features
  • Delivered various live-ops features to enhance game operations
Apr. 2023 – Jun. 2023

Back End Engineer

Froglet Games

Froglet Games is an emerging game development company known for creating engaging and innovative gaming experiences. Focused on backend systems and live-ops features for gaming platforms.

  • Engineered in-game user authentication, leaderboard, and in-game economy systems
  • Managed telemetry systems and Unity Engine upgrades
  • Worked under the mentorship of a former Senior Director of Engineering at Roblox
Jan. 2023 – Apr. 2023

Tools Engineer Intern

Rec Room

Rec Room is a virtual reality multiplayer game that enables users to create and share games and experiences collaboratively. Focused on tooling and UI development to improve developer workflows.

  • Built an in-house build download tool that reduced build download time by 70%
  • Provided UI development and tooling support
Jan. 2022 – Apr. 2022

MLOps Engineer Intern

PointClickCare

PointClickCare is a leading cloud-based healthcare technology provider for long-term and post-acute care facilities. Worked on MLOps pipelines and infrastructure to support machine learning operations.

  • Supported MLOps data pipeline features and infrastructure
  • Created comprehensive test cases for data pipeline validation
  • Monitored and responded to security vulnerabilities
Jul. 2012 – Feb. 2020

Software Engineer

Hubei Xinyegang Co., Ltd.

Hubei Xinyegang is a historic specialty steel producer originally established during WWII. Contributed to the development and maintenance of internal enterprise systems.

  • Handled software submission and maintenance cycles and led system analysis and review
  • Assessed and executed process and system change requests, ensuring necessary data updates
  • Supported external training programs through material creation and coordination
  • Participated in the entire lifecycle of finance and accounting system projects
  • Designed and implemented new modules to expand system capabilities

Education & Certifications

Microsoft

Microsoft

AI Engineer Professional Certificate

Online (via edX)

Sep 2025 – Dec 2025

A professional program focused on building and deploying enterprise-grade AI solutions. The curriculum covers Azure AI services, Machine Learning models, natural language processing, computer vision, and the integration of Azure OpenAI with a focus on responsible and ethical AI development.

Skills Demonstrated

Microsoft Azure AI Azure Machine Learning Azure OpenAI Computer Vision Natural Language Processing (NLP) Conversational AI Responsible AI
Verify Credential
Lakehead University

Lakehead University

Master of Science - MS, Computer Science

Thunder Bay, Ontario, Canada

Sep 2020 – May 2023

GPA: 3.86/4.0

First Class Honours graduate. A consistent volunteer for multiple campus events, clubs, and departments.

Coursework

  • Machine Learning
  • Deep Learning
  • Pattern Recognition
  • Big Data
  • Natural Language Processing

Achievements

  • First Class Honours
  • Campus Event Volunteer
Dalian Minzu University

Dalian Minzu University

Bachelor of Engineering - BE, Software Engineering

Dalian, China

Sep 2008 – Jul 2012

Comprehensive engineering program covering the full software development lifecycle, from mathematical foundations to embedded real-time systems and quality assurance.

Coursework

  • Data Structures and Algorithms
  • Object-Oriented Programming (Java)
  • Software Engineering Integrated Design
  • Operating System
  • Database and Information Management
  • Computer Network
  • Artificial Intelligence Principles
  • Software Quality Assurance and Testing

Achievements

  • CET 4 & CET 6 (English Proficiency)

Projects

CUDA ELM Feature Extraction Benchmark

CUDA ELM Feature Extraction Benchmark

GPU-accelerated Extreme Learning Machine feature-extraction and benchmarking toolkit. Reproducible experiments, modern CUDA/C++ design, and interactive demos for researchers exploring fast ELM variants.

C++ CUDA CMake GoogleTest Google Benchmark Docker FastAPI Python
Aegis AI Gateway

Aegis AI Gateway

A plugin-first AI gateway framework. A small kernel plus seven contracts puts a governed, observable, provider-agnostic pipeline between your apps and any LLM. Built with FastAPI, LangGraph, and Docker.

Python FastAPI LangGraph Docker PostgreSQL
EcoManage

EcoManage

A full-stack renewable energy and resource management dashboard to monitor, analyze, and optimize consumption. Built with React, Node.js, and TypeScript.

TypeScript React Node.js MongoDB
NeighborIQ

NeighborIQ

AI-powered real estate intelligence platform for neighborhood insights, ML price predictions, and automated data ingestion. Built with FastAPI, Vue 3, and Docker.

FastAPI Vue 3 PostgreSQL Elasticsearch
Wu-Tang: Rise of the Deceiver

Wu-Tang: Rise of the Deceiver

An online co-op Action RPG where players team up with the legendary Wu-Tang Clan. Built with Unreal Engine 5, C++, and C#. This game features a robust 3-player multiplayer system set in a unique Afro-surrealist world.

Unreal Engine 5 C++ C# .Net Framework
Bubble UP!

Bubble UP!

A vibrant puzzle game built with Unity, challenging players with fast-paced, Tetris-inspired gameplay. Master chain reactions and aim for high scores in this bubbly, arcade-style adventure.

Unity C# .Net Framework Unity Analytics

Skills

Machine Learning Frameworks

PyTorch
95%
TensorFlow/Keras
90%
Scikit-learn
92%
Hugging Face Transformers
93%

Programming Languages

Python
98%
C++
88%
C#
85%
SQL
85%

AI Specializations

NLP
90%
Computer Vision
85%
Large Language Models
92%
Deep Learning
95%

MLOps & Cloud

Docker/Kubernetes
88%
AWS SageMaker
82%
Azure ML
85%
MLflow
80%

AI Infrastructure

RAG Systems
88%
MCP
85%
Vector Databases
87%
Model Fine-tuning
90%

Latest Blog Posts

The Honest Parts: Residency, Tenancy, and Saying No
May 27, 2026 Ai

The Honest Parts: Residency, Tenancy, and Saying No

You cannot reliably detect where an LLM runs inference. Any product claiming otherwise is overclaiming. This is about the Aegis design decisions that required admitting limits — the residency model built on declared metadata and fail-closed routing, the multi-tenancy deliberately deferred, and why the non-goals list is as load-bearing as the feature list.

Read More
Guardrails That Tell You the Truth About Streaming
May 20, 2026 Ai

Guardrails That Tell You the Truth About Streaming

A guardrail that silently runs at reduced fidelity is worse than no guardrail, because you've stopped watching. The hardest design problem in the Aegis rewrite was keeping safety honest under streaming. Here is the four-verdict model, the capability negotiation that resolved the streaming conflict, and how the same contract ends up governing tool calls.

Read More
Rebuilding an AI Gateway as a Framework, Not a Product
May 13, 2026 Ai

Rebuilding an AI Gateway as a Framework, Not a Product

The first version of Aegis worked, and that was the problem. Every governance feature was hardcoded into one request path, so every change was a core change. This is the story of throwing out the architecture and rebuilding around a small kernel and seven plugin contracts — what that bought, what it cost, and the decisions I'd defend.

Read More