Classifying Indonesia's batik heritage with computer vision.
A full machine-learning product that classifies 20 Nusantara batik motifs, explains the result, and makes the model accessible through a responsive web app.

Informatics graduate · South Jakarta
I'm Mufli Agnal — an AI builder combining machine learning with hands-on LLM fluency to create useful, human-centered products.

01 / Flagship case study
A full machine-learning product that classifies 20 Nusantara batik motifs, explains the result, and makes the model accessible through a responsive web app.

139 of 160 evaluation images classified correctly.
16 matched predictions across 20 internet images.
20 respondents rated the product experience.
All black-box scenarios produced expected results.
Problem framing
→800-image dataset
→Transfer learning
→Model evaluation
→Cloud deployment
02 / Selected builds
Projects where I explore real-time systems, product thinking, and usable interfaces.

A PHP and MySQL storefront with technical product filters, cart, wishlist, payment simulation, and complete customer and admin workflows.
A Python client-server application broadcasting live schedule updates to multiple clients using TCP sockets and threading.
Employee records, attendance tracking, hourly and overtime salary calculation, plus Excel report export.
03 / About
I'm an Informatics graduate from Gunadarma University who loves learning how AI works — then finding a practical way to ship it.
My path started with web, Java, databases, and competitive programming. It grew into machine learning through my thesis: training, evaluating, and deploying an image classifier instead of stopping at a notebook.
I work across ChatGPT, Claude, Kimi, and GLM for research, prompt design, prototyping, debugging, and technical documentation — while verifying outputs and staying aware of model limitations. I'm looking for a team where I can learn fast, contribute carefully, and grow into an AI or machine learning engineer.
04 / Experience
Teaching strengthened the way I communicate technical ideas and support people with different learning speeds.
Taught beginner Java through 10 structured modules and guided 18 students toward completing their practicum projects on time.
Supported 30–40 students per session across Database, Java, and Web Programming practical classes.
05 / Credentials
Verified learning in LLM fluency, software engineering, and competitive programming — backed by hands-on project work.

Claude Academy · Anthropic · Sep 2026
Open certificate ↗
Claude Academy · Anthropic · Sep 2026
Open certificate ↗
Gunadarma I/O · 2023
Open certificate ↗
Gunadarma I/O · 360 minutes
Open certificate ↗
Gunadarma I/O · 450 minutes
Open certificate ↗
06 / Beyond the screen
I enjoy discovering new places, cultures, and ideas. That same curiosity drives how I learn technology: observe closely, ask better questions, and keep exploring.
Have a role, project, or idea?