Selected Work
All projects
26 projects across machine learning, computer vision, OCR, RAG, and full-stack engineering.
ML Intelligent Document Assistant (RAG)
AI / Backend Engineer
RAG-based document Q&A assistant: upload PDF, TXT, or Markdown, then ask questions in Indonesian and get answers with clickable citations. The system refuses to answer when the context is not enough, and general-knowledge answers are clearly labelled so they are never confused with document-sourced ones. Includes per-document private conversations, retrieval quality evaluation, and observability.
+13 
ML Smart Document Scanner & OCR Service
Computer Vision Engineer
Production-ready scanning pipeline: detects a card or document in a photograph, corrects its perspective, enhances legibility, reads it with a neural OCR engine (RapidOCR / PP-OCRv6 on ONNX Runtime), and emits structured JSON where every field carries a confidence and the rule that produced it. The same Dockerfile runs locally, in Compose, and on Railway.
+8 
ML Shale Shaker Cutting Monitoring Web
Full-Stack ML Engineer
Built as my undergraduate thesis (Tugas Akhir): a full migration of the desktop shale-shaker monitoring system to the web. Where v1 used classical OpenCV without deep learning, this version runs deep learning (MobileViT + BiSeNet v2) behind a web tool where users upload a video, mark a 4-point ROI, tune the model, and run live detection with trend visualization.
+4 
ML Wind Turbine Defect Dataset (100K VLM)
Project Lead & QC Specialist
Co-led dataset construction for a vision-language model that spots defects on wind turbine blades: 100,000 real drone-inspection photos across 14 defect classes, from LEP erosion to laminate cracks. With one fellow lead, I ran 5 mentors and 35 annotators across Indonesia: writing inspection guidelines, recording tutorials, holding daily QC syncs, and bridging the company team with the workers on the ground.
ML Shale Shaker Cutting Estimations (Desktop)
Internship Machine Learning Engineer
Desktop-based (v1) computer vision system to monitor and analyze rock cuttings from shale shakers in real-time using CCTV footage, integrated with the WITSML realtime drilling data standard. Built with classical OpenCV image processing, without deep learning. The system was designed to improve the efficiency and accuracy of drilling process monitoring by providing real-time data on rock coverage and composition, thereby supporting operational decision-making in the oil and gas industry.
+2 
ML Tomato Leaf Disease Detection
Machine Learning Engineer
Developed a lightweight deep learning model using TensorFlow Lite to detect diseases in tomato leaves from images. The model, based on transfer learning with MobileNetV2, achieves high accuracy with a compact size, suitable for deployment on resource-constrained devices.
+3 
ML Aerial Image Segmentation
Computer Vision Engineer
Research prototype for my university UAV team: deep-learning segmentation of aerial imagery, satellite and drone, into buildings, roads, and vegetation for geospatial mapping. Built with PyTorch on 500+ annotated images.
ML Prototype Chicken Detection Model
Computer Vision Engineer
Prototype poultry-monitoring system that detects and counts chickens in images and video. Built with Python, Detectron2, and PyTorch around a RetinaNet model with a ResNet-50 FPN backbone, compressed with dynamic quantization and structured pruning.
+2 
ML Road Line Detection
Machine Learning Engineer
This project developed a model to detect road lane markings from image and video inputs at the Samsung Innovation Campus Batch 5 competition. The system was deployed using Streamlit, providing an interactive web interface for users to upload inputs and visualize detected lane lines in image and video processing, with potential applications in navigation and autonomous driving systems.
+1 
Web RigDaily Daily Drilling Report System
Full-Stack Developer
Daily well operations reporting that connects rig and office in one controlled report flow: DDR, 24 hour time breakdown, NPT, cost and AFE, billing, personnel, marine, safety, and rig to office sync. Runs as a Next.js web app for the office and an offline-first Electron desktop .exe with local SQLite for the rig, syncing through an outbox when the network returns. Seven roles from Rig Clerk to Administrator have server-checked permissions, and every approved report locks with a numbered amendment trail.
+11 
Web ShelfOpsRetail Operations Platform
Full-Stack Developer
Retail operations OS that turns CCTV and shelf photos into actionable numbers: stable-ID visitor tracking, per-zone dwell, empty-shelf detection, price mismatch, stock and mutations, procurement, tasks, reports, and a Copilot that answers from SOP and store data with citations. Built as a full Docker stack: Next.js web, Go gateway and worker, Python AI services over gRPC, PostgreSQL 16 with pgvector, fronted by Caddy on a VPS.
+16 
Web MyRouter AI Routing Gateway
Full-Stack Developer
Self-hosted AI routing gateway: one OpenAI-compatible endpoint for many upstream providers (OpenAI, Anthropic, Gemini, DeepSeek, any OpenAI-compatible gateway) with automatic fallback, per-key access limits, and a dashboard. Ships as a single Go binary with the React console embedded and SQLite storage, so there are no external services to operate.
+11 
Web PasFotoIn Browser Photo Editor
Frontend Developer
Browser-based editor for Indonesian ID photos and social crops. Every process runs on the user's device: no uploads, no server, no account. Presets in centimeters (2x3, 3x4, 4x6, 6x9), ICAO-compliant international sizes, automatic subject cutout via an in-browser segmentation model, and an A4 print sheet ready to cut.
+11 
Web LolosCV ATS Resume Builder
Full-Stack Developer
Free resume builder that works backwards from the job posting: paste a LinkedIn, JobStreet, or Glints job URL (or the raw text), let AI break down the requirements, generate a draft, refine it in a drag-and-drop template builder, check the ATS score, then export a pixel-accurate PDF. Runs client-side with a bring-your-own-key AI setup, so resumes, imported PDFs, and API keys never leave the browser.
+5 
Web Visit Bantaragung Tourism Platform
Full-Stack AI Engineer
Built Visit Bantaragung, the official smart-tourism platform for Desa Wisata Bantaragung in Majalengka, as an all-in-one system across 166 commits with the KKN-PPM UGM Simfoni Sindangwangi team. It combines a public Next.js 16 + React 19 landing (destinations, Kampung Herbal B2H catalog, BDB village branding with homestays and MSMEs, BIC information center, Leaflet digital map, gallery), a Git-based content CMS where every save is one commit with zero database, and a RAG chatbot guide with BM25 retrieval, query expansion, reranking, and anti-hallucination guardrails in Indonesian, English, and Javanese.
+10 
IoT Building Automation System Dashboard
IoT Dashboard Engineer
Built the data visualization and integration layer of a Building Automation System monitoring dashboard for the Lab Perancangan Polban. The dashboard unifies WWTP, clean water, and fire protection monitoring over OPC UA, MQTT, and Modbus TCP/IP, with threshold alarms and a simulation mode for testing.
IoT SleepSense IoT Device for Sleep Apnea Monitoring
IoT Engineer
SleepSense is an affordable at-home system for early detection and rehabilitation of Obstructive Sleep Apnea (OSA): an ergonomic belt-thorax sensor records ECG and breathing patterns, a wireless pulse oximeter tracks blood oxygen and heart rate, and a web platform turns both streams into assisted analysis, clinical reports, and a direct line to doctors.
IoT IoT Football Movement Tracker
IoT Engineer
Developed an IoT device to track football players' movement patterns during matches. The device, attached to the back of players' jerseys, records real-time movement data and stores it locally on an SD card in Excel format for subsequent analysis on a dedicated web platform.
DE DataLoom Data Operator Workspace
Full-Stack Data Engineer
Data operator workspace to build, run, and monitor end to end pipelines. Raw data lands in Bronze, is cleaned deterministically into Silver, passes a four rule quality gate, and is promoted to Gold only when fit for use. Every rejected row lands in Quarantine with a reason code, agents only propose while humans approve, and validation rules plus project thresholds are readable in the data contract page.
+18 
DE AI - Sentiment Hub
Data Engineer
Live sentiment dashboard over scraped news on AI, blockchain, finance, forex, and investing. BERT and VADER score each article, React shows the trends, FastAPI serves them, and PostgreSQL with Supabase stores everything, deployed on Render.
+4 
DE Forex Data Monitoring Dashboard
Data Engineer
Dashboard that tracks forex prices as they move and pushes updates through a Telegram bot. Python processes the feed, Grafana visualizes it, and Docker with PostgreSQL holds the pipeline together.
Mobile BIN-GO! Waste Management App
Android Developer
Developed BIN-GO!, a mobile application to address the waste crisis in Yogyakarta. The app integrates Google Maps to locate nearby recycling centers, provides a community forum for waste management discussions, includes a reporting feature for environmental issues, and incorporates gamification to promote sustainable behaviors.
PM Broiler Monitoring v.2
Project Manager
Managed the development of Broiler Monitoring v.2, focusing on enhancing the user interface with a cleaner design and improved functionality. The project involved migrating the frontend framework from Laravel to Next.js while maintaining Laravel for the database API, ensuring a modern and efficient application.
+3 
PM Broiler Monitoring v.1
Project Manager
Led the development of Broiler Monitoring (BroMo) v.1, a web-based information system designed to monitor poultry farms effectively. The application enables real-time monitoring of farm conditions for owners and workers, enhancing operational efficiency.
QA Tokopedia Web Testing
Quality Assurance
Conducted quality assurance testing for the Tokopedia e-commerce platform, focusing on creating and executing three test cases (positive and negative scenarios) to validate the functionality of adding items to the wishlist, adding items to the cart, and performing checkout.
QA BWA Web Testing
Quality Assurance
Conducted quality assurance testing for a web application as part of a course, focusing on creating and executing test cases for two case studies to verify the functionality and reliability of the BuildWithAngga platform.