Ave Syah Shina

•Based in Semarang, Indonesia

Hmm... I'm just an average Information Engineering grad who loves digging in to build digital stuff that leaves a meaningful impression on people. I consider myself a full-stack software engineer, using AI, UI/UX design, and 3D Design to bring my ideas to the community. Check out my About page for the full story.

Download CV

// Blog

My latest writings.

14 — Databases and ORMs — From Raw Drivers to Prisma

A ladder: raw drivers at the bottom (full SQL), query builders in the middle (SQL-shaped JavaScript), ORMs at the top (models, types, migrations). The choice is how much SQL you write versus generate.

Aug 13, 20261 min read

database
typeorm
sequelize
orm
nodejs
mongoose
roadmap:nodejs
prisma
knex
drizzle

01 — What Is AI Engineering? The Role, the Edge, and the Difference from ML

Is an AI Engineer just a rebranded ML Engineer? The boundary that cleared it for me: AI Engineers apply pre-trained models to real problems; training new ones from scratch is someone else's job.

Aug 13, 20261 min read

career
ai-engineering
ml
roadmap:ai-engineer
fundamentals

02 — How LLMs Work: Tokens, Training, Inference, and the Context Window

The next-token trick, unpacked. An LLM is a trained statistical engine: text becomes tokens, a transformer reads them, and the model samples the next token — again and again.

Aug 13, 20261 min read

llm
context
training
inference
tokens
roadmap:ai-engineer
embeddings
fundamentals

04 — Prompt vs Context Engineering: The System Prompt, the Role, the Constraints

'Writing good prompts' bundles two different skills. The split that mattered: prompt engineering crafts the instructions, context engineering builds the whole information environment around the model.

Aug 13, 20261 min read

prompt-engineering
llm
structured-output
system-prompt
roadmap:ai-engineer
context-engineering

03 — Sampling Knobs and Output Rails: Temperature, Top-K, Top-P, and the Rest

Temperature, top-p, max tokens: each knob looks like an isolated setting. The frame that collapsed them: every generation parameter reshapes the same next-token probability distribution before you sample.

Aug 13, 20261 min read

prompt-engineering
llm
temperature
top-p
top-k
roadmap:ai-engineer
sampling
fine-tuning

06 — Types of AI Models: Open, Closed, Pre-trained, and Self-hosted

'Open source good, closed source bad' breaks the moment you pick a real model. The three independent axes that actually decide: open/closed, pre-trained/from-scratch, self-hosted/cloud.

Aug 13, 20261 min read