The story behind the systems
AI/ML Engineer & Data Scientist. 6+ years turning machine learning into production systems that stay running.
Who I Am
AI/ML Engineer and Data Scientist based in Garut, West Java, Indonesia. I work remote-first with clients across Southeast Asia and beyond, building computer vision pipelines, MLOps infrastructure, RAG chatbots, and full-stack ML platforms.
My work spans the full stack of applied ML: from training custom YOLO models for industrial inspection to deploying XGBoost forecasting pipelines with MLflow and Grafana, to building Laravel applications with integrated ML capabilities.
Experience
My career so far, from data engineering through to production ML systems.
- Built automated ML retraining pipelines that run without manual intervention
- Implemented real-time production monitoring with alerting and anomaly detection
- Designed multi-stage computer vision inference that exceeds single-stage baselines
- Owned the full MLOps lifecycle: experiment tracking, versioning, serving, and observability
- Delivered end-to-end financial systems with double-entry accounting and audit logging
- Analysed digital media performance across platforms for public sector communications teams
- Built dashboards and reports that made engagement data actionable for non-technical decision-makers
- Ran exploratory data analysis to surface trends and inform campaign strategy
- Supported migration from legacy file systems to SharePoint, including metadata standardisation and data consistency validation
I also contribute to open source projects, including provider modules for Indonesia's largest job portals.
See my open source work →My Philosophy
"Every system I ship includes monitoring, evaluation pipelines, and clean architecture. Getting the model right is only one part of it."
The gap between a Jupyter notebook and a production system is where most ML projects fail. My approach bridges that gap:
- Ship early, measure everything: models without evaluation metrics are just opinions
- Clean architecture over clever models: a simple model with a solid pipeline beats a fancy model nobody can deploy
- Document the "why": architecture decisions, tradeoffs, and lessons learned are as valuable as the code
- Build for the person after you: often that person is you, six months later
The NDA Reality
Most of my best work is under non-disclosure agreements with enterprise clients. You won't find the production code on GitHub. Where I do have public repos, they're intentionally simplified demonstrations. The production versions had more features, better monitoring, and were built for real institutional use. The case studies on this site explain the engineering thinking behind each system.
If you're a recruiter or potential client: the NDA badge on a project means I built a more sophisticated version for a paying customer. The one exception is Sovereign Ledger, built for KPRI Warga Kesehatan Kabupaten Garut, deployed on-premise, and in active daily use. I'm happy to discuss any of these in detail.
What I'm Looking For
I'm actively evaluating opportunities for ML Engineer, Data Scientist, and AI/ML roles, remote or hybrid. I thrive in environments where:
- ML is treated as an engineering discipline, not a research experiment
- There's room to own the full pipeline, from data to deployment
- The team values clear communication and documented decisions
- Production reliability matters as much as model accuracy
Beyond Work
When I'm not building ML systems, I'm probably reading about them, or writing about them. I maintain this blog as a way to clarify my thinking and contribute back to the community. I also enjoy chess (hence the Magic Chess project), exploring system design patterns, and mentoring junior engineers breaking into ML.
Education
Bachelor of Computer Science, Universitas Terbuka (GPA: 3.36 / 4.0)
Transferred from UIN Syarif Hidayatullah Jakarta to pursue distance learning while building full-time professional experience in parallel.
Certifications
- Python – Data Science, SanberCode (2020)
- Certified International Specialist Data Modelling (CISDM), Cybertrend (2019)
- Certified International Project Manager Associate (CIPMA), Cybertrend (2019)
- Certified International Supply Chain Associate (CISCA), PASAS Institute (2017)
- Microsoft Technology Associate (MTA), Microsoft (2018)