AI Engineer • Computer Vision • Efficient AI & Edge Deployment
_{Backed by Software Engineering, Networking, Cybersecurity & AIoT}

I build systems at the intersection of Computer Vision, Applied AI, and Software Engineering — from training and evaluating models to shipping the APIs, dashboards, and infrastructure that put them in front of users. My main technical interest is efficient, real-time object detection and edge deployment; my day-to-day shipped work is full-stack AI applications (FastAPI + Next.js), local-first AI tooling, and security utilities.

That work is backed by a full-stack engineering background (FastAPI, Laravel, Next.js, Flask) and a growing foundation in Linux systems, networking, IoT, and cybersecurity tooling — because a model is only useful once it's running inside a real system: an API, a database, a server, a network, a device.

MODEL → CODE → API → DATABASE → SERVER → NETWORK → SECURITY → EDGE DEVICE

I'm an Informatics Engineering student at Universitas Muhammadiyah Riau, currently deepening applied AI research and open-source tooling that connects model development to the systems AI actually runs on.

AI            → Object detection (YOLO), transfer learning, model evaluation, Grad-CAM
Engineering   → FastAPI model-serving, Next.js frontends, Redis/RQ workers, Docker, GitHub Actions CI
Deployment    → Local-first / self-hosted AI apps; ONNX / TensorRT / edge inference (developing)
Systems       → Linux, networking fundamentals, AIoT, server dashboards & monitoring
Security      → Python security utilities, reverse-engineering tooling, coordinated-messaging analysis
Research      → Reproducible experiments, architecture comparison, benchmarking

Shipping / recently shipped

🐝  Stingers — 100% local AI chatbot: streaming SSE, 119 models via 9Router, RAG & memory
🌉  Bifrost-gateway — self-hosted LLM routing gateway (OpenAI + Anthropic-compatible endpoints)
🔭  OpenReframe — AI media reframing platform; core image/video pipeline working end-to-end
🌿  AgroVision-AI — leaf-photo plant disease diagnosis (working MVP, 88.5% test accuracy)
🌾  CropYield-AI — crop yield prediction from climate & land data (working MVP, R² 0.984)
🌦️  Weather Risk Prediction — Sumatera (NASA POWER + NOAA, Prophet + SVM)
🩺  Retinal OCT Classification — ResNet50 vs EfficientNetV2S vs DenseNet121 with Grad-CAM
🗄️  Mneme — local-first Markdown knowledge vault (Rust/Tauri + React)
🌐  ProtoKarya Hub & 🏢 Creativ Potlot — full-stack client/competition projects (Next.js)

CORE
Python · Computer Vision · FastAPI · Next.js
TensorFlow/Keras · scikit-learn · Git/GitHub

APPLIED
Object Detection (YOLO + OpenCV) · Transfer Learning · Grad-CAM · Model Evaluation
Redis/RQ workers · Docker · REST API · Laravel · Flask · Linux
LLM Chat Apps (SSE, RAG, Memory) · LLM Gateways · Security Utilities in Python

DEVELOPING
PyTorch · ONNX · TensorRT · Quantization · Pruning · Knowledge Distillation
MLflow · MLOps · Edge AI · GPU Inference
Networking Fundamentals · AIoT (ESP32/Jetson/Raspberry Pi/MQTT)

Efficient object detection · model compression (quantization, pruning, distillation) · real-time computer vision · edge AI · explainable AI · AIoT · reproducible ML experiments · model benchmarking

Supporting: AI systems infrastructure · applied LLM/RAG applications · network-aware application design · cybersecurity automation

AI / ML

Python TensorFlow Keras PyTorch (developing) Ultralytics/YOLO OpenCV Scikit-learn Prophet NumPy Pandas Matplotlib

AI Optimization / Deployment

Docker FastAPI Redis / RQ GitHub Actions Linux ONNX (developing) TensorRT (developing) MLflow (developing) GPU Inference (developing)

Applied AI / LLM Apps

Streaming SSE RAG Vector Embeddings LLM Gateways (OpenAI / Anthropic-compatible) Long-term Memory

Programming

Python PHP JavaScript TypeScript Dart Rust (via Mneme) Bash SQL

Backend & Database

FastAPI Flask Laravel Node.js / Express PostgreSQL MySQL SQLite SQLAlchemy Prisma Supabase

Frontend

Networking

TCP/IP HTTP/HTTPS REST MQTT Linux Networking VPS

Cybersecurity

Reverse Engineering Deobfuscation Python Security Scripting Network Scanning Server Hardening

IoT / Embedded

ESP32 Arduino LoRa MQTT Raspberry Pi (developing) NVIDIA Jetson (developing)

Mobile

DevOps / Tools

AI / Computer Vision

Developer Tools & Infrastructure

Software Engineering

Cybersecurity

Other Small Projects

Forks I work from for upstream contributions and study: freeCodeCamp, maigret, local-deep-research, Octop, open-code-review, claude-mem, uptime-kuma, excalidraw, tiptap.

Recent contributions to external open-source projects:

_{Table kept manually — update the status column as PRs get reviewed/merged.}

Claude 101 Anthropic	Introduction to Claude Cowork Anthropic	Guide to Learn Python with AI DQLab	Guide to Learn SQL with AI DQLab	Guide to Learn R with AI DQLab
Intro to Data Science with Python DQLab	Intro to Data Science with R DQLab	Data Analyst with Excel DQLab	Data Analyst with SQL & Python DQLab	Sertifikat Lainnya Entrepreneurship Award
ProtoKarya Liga Talenta Mahasiswa Indonesia — LLDIKTI XVII	Juara 2 — Business Poster & Pitching Liga Talenta Mahasiswa Indonesia — LLDIKTI XVII

• Build efficient, real-time computer vision systems and deploy them beyond prototypes
• Take model optimization (quantization, pruning, distillation) from experimentation to edge deployment
• Develop end-to-end, reproducible ML pipelines (training → serving → monitoring)
• Publish OpenReframe as a stable, documented open-source release (SAM masks, outpainting)
• Expand AgroVision-AI to the full 38-class PlantVillage dataset with a CNN backend, and evolve toward field data such as oil palm
• Extend Stingers with tool calling, an artifact canvas, and local voice input
• Strengthen networking and systems engineering fundamentals
• Grow a library of ethical hacking & security tooling for authorized research
• Publish technical write-ups on applied, reproducible AI research
• Continue contributing to established open-source projects (freeCodeCamp, maigret) alongside personal projects

"I don't just build models — I build the systems they run on."

_{Muhammad Almas Albirra Hamid · Informatics Engineering, Universitas Muhammadiyah Riau · Indonesia}