Hire the Best Computer Vision Engineers
in Indonesia

Training the model is the easy part. The real challenge begins when the prototype has to be deployed. My philosophy is straightforward: I don't stop when the model works; I keep going until the solution works. Commercial projects taught me that clients rarely remember a model's accuracy—they remember whether the solution solved their problem. 1. What I build ▸ Computer Vision: object detection, image classification, face recognition, tracking, segmentation, OCR, anomaly detection ▸ Intelligent Surveillance: face recognition, access control, CCTV analytics, people counting, event detection, real-time monitoring ▸ AI Deployment & Integration: MQTT, REST APIs, Docker, FastAPI, real-time streaming, event-driven architectures ▸ Edge AI: NVIDIA Jetson, Raspberry Pi, TensorRT, ONNX, optimized real-time inference ▸ Data Science & Machine Learning: predictive modeling, feature engineering, data preprocessing, model evaluation, statistical analysis, SQL ▸ AI Applications: intelligent inspection, agricultural AI, industrial vision, workflow automation ▸ Core Technologies: Python, OpenCV, TensorFlow, PyTorch, Scikit-learn, Pandas, NumPy, SQL 2. Real work I have shipped • Real-time Face Recognition & Access Control Developed in collaboration with Milesight and Fartak, delivering a real-time, multi-camera face recognition platform capable of processing live IP camera streams with automated access control through MQTT. • Intelligent Irrigation Management Platform Designed for Fartak and a metropolitan municipality, integrating ESP32 IoT sensors, environmental monitoring, and machine learning to automate irrigation planning and optimize water usage across a 30*20 km green space. • Smart Vineyard Monitoring Platform Developed a computer vision solution for a smart vineyard, analyzing camera feeds to detect grapevine diseases, map disease distribution across the farm, and provide actionable insights for early intervention and crop management. 3. How I Work ✓ Clarify requirements upfront to eliminate misunderstandings before development begins ✓ Communicate proactively with regular progress updates ✓ Adapt quickly to changing requirements while keeping the project on track ✓ Every architectural decision is made with long-term scalability in mind—not just today's requirements. ✓ Optimize for business impact before technical complexity. ✓ Treat every project as a long-term product, not a one-time delivery. Every successful project starts with a clear understanding of the problem. Whether you're building a new computer vision system or improving an existing one, I'd be glad to discuss your project and see if we're the right fit.

I love building solution for computer vision problems, implementing machine learning methods, developing efficient program to solve complex tasks, and of course, making things in the world of electronics. I am feeling extremely passionate on details, with senses of art . For more than 10 years, computer programming has already been in my "list of routine activity". I am currently working to realize commercial-based smart vision system.

Experienced in building real-time applications, defect detection systems, and interactive user interfaces using Python, GLCM, SVM, and Tkinter. Skilled in creating efficient digital tools and web applications with HTML, CSS, Google Apps Script, and Firebase. Strong focus on experimental projects, with a commitment to advancing expertise and building a professional portfolio in AI and technology.

I am a fresh graduate with a Bachelor of Computer Engineering at Institut Teknologi Sepuluh Nopember (ITS), Surabaya, Indonesia with a very satisfactory grade. I've also participated in the Bangkit 2023 program as a Cloud Computing Cohort which is a program held by Google and delivered with support from industry experts from GoTo and Traveloka. I have a keen interest in learning about computer vision especially about YOLO (You Only Look Once) and my thesis is about changing the backbone of YOLOv8 by upscaling it to increase the mAP (Mean Average Precision) as part of the accuracy metrics. The title of my thesis is "Deteksi Lumba - Lumba Putih China Menggunakan YOLO" or "Detection Of Chinese White Dolphins Using YOLO" in english. The dataset is pretty challenging because of how similar the objects are among the classes. • Proficient in YOLO (You Only Look Once), Pytorch, OpenCV • Knows Python, HTML and CSS • Skilled in IOT modules such as Arduino and ESP32

👨‍🎓 About me • Graduated in Computer Science, Machine Learning, Deep Learning (additional education at AI university and Deep School and many more) • More than 6 years of experience in ML / DL / CV / GenAI products development • Full cycle - embedding software integrated with AI techs . Now I work with all stages of DL production: from creation of model (or even gathering a dataset) up to deploy the product to the production • Optionally: a team of different skills engineers and we can perform any complicated tasks for you • All conditions are negotiable 🎯 MAIN STREAMS: Computer Vision: • 3D computer vision (3D segmentation, SLAM, SfM, NeRF, etc.) • Object detection • Semantic or instance / panoptic segmentation • Object tracking and object reidentification • Face detection and face recognition • Optical character recognition (OCR) • Images matching • Images classification Generative AI: • Generation of a consistent character • Outfit and pose swap for generated character • Face swap, facial expressions • ComfyUI API integration • Train LORA, DreamBooth for Stable Diffusion 1.5, SDXL, FLUX MLOps Data scraping Data visualization 🛠️ MAIN TOOLS: • ML/DL: PyTorch, Keras / Tensorflow, ScikitLearn, XGBoost, and many more • 3DML: Open3d, PyTorch3d, PointNet, Blender (points cloud sampling), NeRF • Experiment management: ClearML, W&B • Generative AI: FLUX, SDXL, Stable Diffusion 1.5, ComfyUI, ControlNet, IPAdapter, Automatic1111, KohyaSS, etc • Data processing: Numpy, Pandas and many more • Data visualization: Grafana, Matplotlib, Plotly • Data scraping: Selenium, BeautifulSoup • Image processing: OpenCV, PIL, ScikitImage • Inference frameworks: TensorRT, ONNXRuntime, OpenVino, Triton Inference Server • Dev/MLOps: Git, Docker, Docker-Compose, RabbitMQ, Ansible, Flask, FastAPI, Ubuntu, Bash, DVC, GitLFS • Server monitoring: Prometheus, Grafana, Portainer and many more 💼 MY PROJECTS: • Cars segmentation, generation shadows and reflection maps with SDXL, view classification, 360-spin • 3D Segmentation with point clouds (Pointnet2) for jewelry industry • Customers traffic counting and fraud detection for retail • Counting visitors at shops using video from survaillance cameras • GameDev generation maps from depth maps using Stable Diffusion and ControlNet • Generation of AI influencers, swapping clothes, pose application, swapping faces and facial expressions • Swapping of floor texture at indoor photos (searching for semantic lines, searching for vanishing points, texture scaling, shadow/light applying, floor segmentation, texture swap, reflection map generation, production) • Audio classification for babies' cry • Medical data segmentation and classification • Car plates detection and recognition • Detecting neural networks for wholesales • Many other projects

? AI & COMPUTER VISION SPECIALIST | YOLOv8 | DEEP LEARNING | PYTHON I'm a recent physics graduate with a focus on computational physics and scientific data analysis. I specialize in building deep learning models for object detection using YOLOv8, with experience applying oriented bounding boxes (OBB) for scientific track visualization in cloud chamber experiments. My core strengths include: • Developing custom object detection pipelines using Python and YOLOv8 • Hands-on dataset creation and annotation with Roboflow • Performance evaluation with mAP, precision-recall, F1 curve, and confusion matrix • End-to-end project execution using Google Colab and visualization tools If you're looking for someone with a strong research background, proven AI project experience, and the ability to communicate technical work clearly, let's work together.