Skip to content
View HuaTNA's full-sized avatar
  • University of Toronto

Block or report HuaTNA

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
HuaTNA/README.md

Hey, I'm Hua 👋

I build LLM applications, retrieval systems, and cloud automation that turn everyday workflows into useful software.

🎓 Industrial Engineering at the University of Toronto, with minors in AI and Engineering Business · Expected May 2027
📍 Toronto, Canada 🇨🇦

🔭 What I've been building

  • Career Workspace — matches a structured personal profile to job descriptions, drafts tailored resumes and cover letters, and exports PDFs. Built with Next.js, FastAPI, Claude API, Supabase, and LaTeX.
  • Clinical Tutor — an AI tutor for practicing clinical reasoning with 2,000 de-identified ICU cases, combining knowledge graphs and semantic retrieval with guided questions and feedback.
  • Fraud Detection MLOps — a reproducible training-to-inference workflow with real-time and batch predictions, FastAPI, MLflow, Kafka, and Docker.

💼 What I've worked on

As an AI Application Engineer Intern at Spark Growth, I built internal AI tools and automation for a growth agency serving DTC brands:

  • AI agents & retrieval — Python agents on Cloud Run with 31 API tools, shared memory components, and permission-aware retrieval across meeting, email, and Slack sources.
  • Meeting-to-Task Automation — a pipeline that extracts action items and creates tasks in Productive CRM, processing 100+ meetings weekly with 95% project classification accuracy.
  • Creative Intelligence Agent — natural-language analysis of Instagram, Facebook, and LinkedIn creative performance using Claude tool use, FastAPI, and Cloud SQL.
  • CRM & reporting automation — webhook integrations, data pipelines, and a full-stack report builder with web, PDF, and PowerPoint exports.

🌱 Currently exploring

  • RAG architecture, knowledge graphs, and agent memory
  • Agent evaluation, observability, and reliable tool use
  • Data Structures & Algorithms — interview prep continues 🫠
  • French 🇫🇷 — keeping the Duolingo streak alive

🛠 Tech stack

Languages     Python · TypeScript · SQL · JavaScript
AI / ML       Claude API · Vertex AI · RAG · Embeddings · PyTorch
Backend       FastAPI · Flask · REST APIs · OAuth 2.0
Frontend      Next.js · React
Data          Cloud SQL (MySQL) · PostgreSQL · Supabase · ChromaDB
Cloud         GCP · Cloud Run · Cloud Functions · Docker
Quality       pytest · Vitest · GitHub Actions · Agent evaluations
Automation    Google Apps Script · Webhooks · Zapier

💞️ Let's build something

I'm interested in collaborating on LLM applications, AI agents, retrieval systems, and data pipelines — especially tools that automate away the boring stuff.

Open to AI / LLM Application Engineer and ML Engineer opportunities.

Outside the terminal: landscape photography, fingerstyle and classical guitar, and learning French.

📫 Reach me

Email · LinkedIn · GitHub

Popular repositories Loading

  1. healthcare_ai healthcare_ai Public

    TypeScript 3

  2. MBTI-Prediction- MBTI-Prediction- Public

    Jupyter Notebook 2

  3. Auto-Resume Auto-Resume Public

    Python 2

  4. HuaTNA HuaTNA Public

    Config files for my GitHub profile.

    1

  5. fraud-detection-mlops fraud-detection-mlops Public

    Real-time fraud detection with ML model deployment, Kafka stream simulation, and monitoring

    Python 1

  6. WARP_Shoe_company WARP_Shoe_company Public

    AMPL