Open to remote roles · UTC+8
I'm Hasan Basri, an IT systems administrator in Indonesia. For over a decade I have kept networks, servers and academic information systems running for two organisations at once, including four national-scale government reporting platforms where a missed deadline has real consequences. I am now completing a Master's in Computer Systems, with thesis research comparing CNN backbones for a face-recognition attendance system that runs in production.
Systems Administrator & CV Engineer
$ candidate.status()
> availability: open_to_work
> stack: ["Python", "PyTorch", "Linux", "Networking", "SQL"]
> research: "CNN backbone comparison, face recognition"
> overlap: "Asia-Pacific full day, Europe mornings"
Professional Summary
Proven operational reliability, combined with machine learning research I am taking from paper to production.
I look after the part of IT that is not allowed to fail: servers, networks, databases and statutory reporting. At Perumda Air Minum Kolaka Utara, a regional water utility, I run the operational systems and network. At Universitas Muhammadiyah Kolaka Utara I administer the academic information system and own the institution's submissions to four national platforms: PDDikti, SISTER, SINTA and KIP Kuliah. These are the Indonesian government's higher-education data systems, where records must reconcile exactly and deadlines are fixed by the ministry.
Over the last few years I have moved deliberately toward computer vision and deep learning. My thesis compares several CNN backbone architectures for a face-recognition attendance system: I collected the dataset myself under real lighting conditions, trained and benchmarked the models, and measured both accuracy and inference time on the hardware actually available on site.
That combination is the reason I am worth talking to. I am not a researcher who stops at a notebook, and not an administrator who is uncomfortable with code. I think about deployment, database integration, access control and who maintains the thing in two years, because those are already my job.
New technology is only useful if somebody can keep it running every day without drama.
Evidence
Things I have built or run, rather than adjectives about myself.
Master's thesis · 2025 to 2026
A CNN-based attendance system for a regional water utility, replacing manual sign-in sheets. I built the dataset from scratch across varied lighting and camera angles, benchmarked multiple backbone architectures for the accuracy and inference-speed trade-off, and wired the chosen model into the attendance database with handling for undetected faces.
Code on GitHubOngoing · Universitas Muhammadiyah Kolaka Utara
I own the university's reporting into four Indonesian government platforms. That means reconciling student, lecturer and course records across systems that disagree with each other, designing consistent naming and coding schemes so records match, and correcting historical data before each semester's submission window closes.
Internal government systems, not publicly linkable.
Since 2017 · YouTube and Facebook
I produce short technical explainers on AI tooling, digital productivity and computer fundamentals for an Indonesian audience. It keeps me current, and it is the reason I can explain a system design to someone who does not share my background, which turns out to matter more in distributed teams than any individual tool on my CV.
Watch the channelCapabilities
Grouped by the work they belong to, not scored as percentages, so you can match them against a role.
Face recognition, image preprocessing, dataset collection and augmentation, training and comparing CNN architectures, and evaluating models against latency budgets.
Automation scripts for data processing, dataset manipulation, analysis of experiment results, and integration with institutional databases.
Academic data lifecycle management, statutory reporting pipelines, data migration, and reconciliation across systems that do not agree.
Router and switch configuration, network segmentation, on-premise server maintenance, service monitoring, and incident response.
Operating system deployment and recovery, BIOS and storage controller configuration, hardware diagnosis and repair across a mixed fleet.
Responsive interfaces built without framework overhead, institutional website administration, page performance work, and technical SEO.
Track Record
Regional water utility, Kolaka Utara, Indonesia
Operator for PDDikti, SISTER, SINTA and KIP Kuliah
Master's thesis, Computer Systems
YouTube and Facebook
Producing educational content on AI tools, digital productivity and computing fundamentals. The practice of compressing a technical idea into something a stranger can follow transfers directly to documentation and cross-functional communication.
View channelQualifications
Universitas Handayani Makassar
Universitas Terbuka
Systems analysis, database management, and IT infrastructure governance.
How I Work
A short walk-through of the process I follow, from defining the problem through to handing over something the team can maintain without me.
// Press the button to play the sequence.
> engineer: Hasan Basri
> case: face-recognition employee attendance
Contact
Open to full-time roles, contract work and research collaboration. I usually reply within 24 hours on working days.
Based in Indonesia on UTC+8. That is a full working-day overlap with Singapore, Hong Kong, Perth and Manila, and a comfortable morning overlap with Europe. Open to remote, hybrid and on-site arrangements.