Present - 2026
Student at HRD Center
Main course: Mastering Local LLM. Learning how to run, evaluate, and build useful tools around local language models.
Hi, I am Limhai , a student interested in local AI, backend development, and turning difficult ideas into small things people can actually use. I enjoy learning by building, breaking, fixing, and improving one step at a time.
Currently focused on Local LLM workflows, Spring Boot APIs, and clean interfaces for practical learning tools.
Projects
Prompt Engineering & LLM Research
Exploring how to design effective prompts, manage context, and build reliable LLM-powered applications through practical experiments and research.
Key learning areas
01
Prompt Engineering
Designing structured prompts for better LLM responses.
02
Context Engineering
Managing context and relevant information effectively.
03
Prompt & Context Compression
Token optimization and LLMLingua.
04
RAG vs. Fine-Tuning vs. Long-Context
Understanding approaches, trade-offs, and use cases.
05
Retrieval-Augmented Generation
Grounding LLM responses with retrieved information.
06
LLM Fundamentals
Hallucinations, embeddings, vector databases, and model limitations.
Hands-on experience
- Experimenting with local LLMs using Ollama.
- Building small RAG projects using Python.
- Exploring embeddings, semantic search, and vector databases.
- Learning prompt optimization and retrieval techniques.
Journeys
Sub subject
Spring Boot
Practicing backend fundamentals with Java, REST APIs, layered architecture, and clean service design.
Focus
AI + Backend Development
Interested in combining local AI, reliable APIs, and simple interfaces that help people learn and work faster.
Learning Notes
Short, practical notes from my current study path: local AI, backend APIs, and the small decisions that make a project easier to understand.
Running Local LLMs
Notes on model setup, prompts, inference speed, and keeping AI workflows private on a local machine.
Spring Boot API Design
Practical patterns for controllers, services, DTOs, validation, and predictable backend structure.
RAG Basics
A beginner-friendly breakdown of chunking, embeddings, retrieval, and grounding answers with documents.
Open to learning and collaboration