Skip to content
View FlosMume's full-sized avatar

Block or report FlosMume

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
FlosMume/README.md

Samuel Huang

Machine Learning Engineer | LLM & Agentic AI | RAG | Applied ML | Engineering AI

I build practical AI systems spanning agentic workflows, retrieval-augmented generation, machine learning, GPU computing, engineering applications, and AI education.

Featured Projects

🤖 Agentic AI

CoursePilot

An agentic course-development and quality-assurance system built with Strands Agents SDK and AWS.

  • Requirements extraction and constraint analysis
  • Agent-based course planning
  • Deterministic validation
  • Instructor-in-the-loop review
  • AWS Bedrock integration

Multi-Agent Course Intelligence

A comparative implementation of course-intelligence workflows using native Python and multiple agent frameworks.

Areas explored include:

  • Agent roles and orchestration
  • Research and summarization
  • Validation
  • Multi-agent workflow design
  • Framework comparison

🧠 LLM Systems

CareMind

An end-to-end retrieval-augmented generation application using LLMs, vector retrieval, and contextual evidence.

Efficient LLM Fine-Tuning

Experiments with Llama fine-tuning using QLoRA and Unsloth, with emphasis on efficient model adaptation.

⚙️ Engineering AI

AI in Engineering

Applications of AI and machine learning to engineering problems, including predictive maintenance, computer vision, structural health monitoring, HVAC optimization, and edge/cloud deployment.

🚀 GPU Computing & AI Acceleration

Hands-on work with:

  • CUDA C/C++
  • GPU memory management
  • CUDA streams
  • Thrust
  • GPU image processing
  • ONNX
  • TensorRT
  • AI inference acceleration

📊 Data & Machine Learning Engineering

Additional work includes:

  • Databricks
  • Data pipelines
  • Machine learning
  • XGBoost
  • Statistical analysis
  • Data visualization
  • Reproducible analytics

Teaching & Course Development

I develop and teach university-level courses and learning materials in areas including:

  • Artificial Intelligence
  • Data Science and Analytics
  • Object-Oriented Programming
  • Machine Learning
  • LLM and Agentic AI
  • GPU Computing
  • Engineering applications of AI

Current Interests

Agentic AI · Multi-Agent Systems · LLM Evaluation · RAG · AI Engineering · GPU Computing


Thanks for visiting my GitHub profile.

Pinned Loading

  1. CoursePilot-Agent CoursePilot-Agent Public

    An agentic course-development and quality-assurance system built with Strands Agents SDK, Amazon Bedrock, and deterministic validation.

    Python

  2. multi-agent-course-intelligence multi-agent-course-intelligence Public

    A comparative multi-agent course-intelligence system exploring native Python and agent frameworks for research, summarization, validation, and workflow orchestration.

    Python

  3. CareMind-Streamlit CareMind-Streamlit Public

    End-to-end RAG application using Streamlit, ChromaDB, LLM-based retrieval, a SQLite drug database, and contextual evidence reasoning.

    Python

  4. LLAMA-qLoRA-Unsloth-Starter LLAMA-qLoRA-Unsloth-Starter Public

    Efficient Llama fine-tuning with QLoRA and Unsloth for supervised instruction tuning and model adaptation.

    Python 1 1

  5. AI-in-Engineering-Demo AI-in-Engineering-Demo Public

    Applied AI for engineering: predictive maintenance, computer vision, SHM, HVAC optimization, and edge/cloud deployment.

    Python 1

  6. CUDA-AI-Inference-Starter CUDA-AI-Inference-Starter Public

    GPU-accelerated AI inference with CUDA, ONNX, and TensorRT, including benchmarking and deployment-oriented examples.

    Cuda 1