What we have built, and what we have worked on.

Issue and follow-up portal

An interdepartmental issue and complaint system for organisations that run several departments. Anyone raises an issue against a department; its supervisor and manager are notified; the supervisor assigns the work; and the manager verifies the closure — deliberately never the supervisor who handed it out.

Roles are held per department, so one person can be a supervisor in one and a manager in another. Configurable SLA targets, a role-aware dashboard, attachments, Excel export, email notifications, Google sign-in and an audit trail, in a bilingual English and Indonesian interface.

Outcome areas: workflow and approval systems, role-based access control, operational reporting, notification and escalation.

Warehouse and order fulfilment platform

A platform giving a distribution and warehousing business real-time control over fulfilment, replacing spreadsheets and status calls with one searchable source of truth. Every invoice is timestamped through warehouse intake, picking, checking, team-lead approval, warehouse admin, reception and finance sign-off.

Each transition is attributed to whoever made it, so management sees how long a stage takes and who handled it. Staff are routed by role to a workspace built for their job, inventory is tracked by location with shipping-receipt validation, and every view is searchable with one-click Excel export.

Outcome areas: fulfilment tracking, role-based access control, inventory and stock movement, operational reporting, REST API integration.

Generative AI and document intelligence

Generative-AI solutions for document-driven and human-centred applications, including Vision-Language Models and multimodal transformers for information extraction, semantic reasoning, and image-text understanding, plus a Retrieval-Augmented Generation system for document processing.

Abdul · Generative AI · Vision-Language Models · document intelligence · information extraction · semantic search · RAG systems

Medical image analysis and reconstruction

Deep-learning and transformer-based systems for detecting multiple abnormalities from medical images: complex visual data, multi-label detection, and AI-assisted decision support.

Alongside it, cone-beam CT reconstruction for dental and maxillofacial imaging — detecting patient motion during a scan and compensating for it without an iterative solver, so a scan degraded by movement can still be reconstructed.

Anneke and Abdul · Deep learning · transformer-based models · image reconstruction · motion compensation · decision support

Aerial image and map matching

Computer-vision research involving aerial imagery, geographic data, image alignment, and map matching — the groundwork behind inspection, mapping, and location-based operational systems.

Anneke and Abdul · Aerial-image processing · geospatial analysis · image alignment · visual matching

Machine-learning forecasting

A machine-learning application for rainfall forecasting in Indonesia, covering predictive modelling, data analysis, and practical data-driven application development.

Anneke · Predictive modelling · forecasting · data analysis