Computer Vision · NLP · Time-Series · Reproducible ML
I'm Budi Juarto, an AI engineer, researcher, and Computer Science lecturer at BINUS University Bandung, Indonesia.
I work across computer vision, natural language processing, and time-series modeling, connecting research experiments with usable applications. My interests range from image recognition and on-device inference to sentiment-aware forecasting and recommendation systems.
My engineering priorities are honest evaluation, reproducible experiments, and maintainable deployment. I care about the complete system: the data, the model, the interface, and how it runs outside the development environment.
Research informs the model. Engineering makes it usable.
A selection of my applied AI, research, and software projects. Each repository provides its own implementation details and setup instructions where available.
|
EDGE AI · MOBILE APPLICATION On-device urban issue classification in a Flutter reporting app, supported by an image-model training pipeline and a reproducible Docker environment.
|
NLP · TIME-SERIES · ANALYTICS A BBCA forecasting research pipeline combining price data, macroeconomic indicators, and news sentiment, with model comparisons and an analytics dashboard.
|
|
COMPUTER VISION · MODEL BENCHMARKING Image classification research comparing CNN, deformable convolution, and hybrid attention architectures across 31 Indonesian herbal plant classes.
|
COMPUTER VISION · CULTURAL HERITAGE A batik image dataset, experiment notebooks, and a GUI for exploring image classification with CNN, Vision Transformer, and YOLO models.
|
|
RECOMMENDATION SYSTEMS · FULL-STACK Ingredient-aware recipe recommendations using weighted content-based matching, delivered through a full-stack Next.js application.
|
CLOUD-NATIVE · MONITORING A hands-on learning lab connecting FastAPI metrics to Prometheus and Grafana through Docker Compose, with logging and alerting examples.
|
| Languages | Python · C++ · JavaScript · TypeScript · Go |
| Machine Learning | PyTorch · TensorFlow · scikit-learn · Hugging Face · OpenCV · YOLO |
| Applications & APIs | FastAPI · Flask · Streamlit · Next.js · Flutter |
| Data & Infrastructure | PostgreSQL · MySQL · Linux · Docker · Git · GitHub Actions |
Define the problem → Audit the data → Build a baseline → Evaluate carefully → Deploy → Iterate
Evaluation before claims. Choose metrics that fit the task, check for data leakage, and compare against meaningful baselines.
Reproducibility before convenience. Keep environments, configurations, and experiment artifacts organized so results can be revisited and verified.
Deployment as part of the design. Consider inference latency, resource usage, API design, logging, and maintainability—not only model accuracy.
Alongside development, I teach algorithms, operating systems, and cloud computing, and supervise student projects connecting AI research with practical software development.
I'm interested in collaborations around applied computer vision, lightweight inference, NLP, time-series analysis, and reproducible machine learning.
| Language | Proficiency |
|---|---|
| Indonesian | Native |
| English | Professional working proficiency |
| Japanese | JLPT N2 — Passed |
Interested in AI/ML engineering opportunities in Japan and international research collaborations.
View contribution activity
This chart is provided by a third-party service. Repository links remain available if the chart is temporarily unavailable.
Have an applied AI problem, a research idea, or an engineering opportunity to discuss?
Email: budi.juarto@gmail.com
Portfolio: budijuarto.com
GitHub: @SeedFlora
Build thoughtfully. Evaluate honestly. Make it reproducible.