AI & machine learning
Models trained and evaluated on your data — classification, detection, anomaly and forecasting — with the failure cases written down, not hidden.
Software
Software, AI & image processing
Software here is not a side product of the hardware. We build cross-platform applications, machine-learning models evaluated on your own data, and image-processing pipelines precise enough to stand next to a gauge reading.
Working with
- Python
- OpenCV
- NumPy
- Flutter
- PostgreSQL
- Supabase
- Docker
- GUM
What we do
Image processing & measurement
Sub-pixel measurement, multi-view fusion, calibration and uncertainty analysis to GUM — numbers you can defend in a QC record.
Cross-platform applications
One codebase for web, Android and iOS. Offline-first where connectivity drops, which in the field it does.
Backends, data & integration
Databases with real access control, APIs, dashboards, and integration with the systems you already run.
How a software project runs
This is the process for this discipline. The other three run differently — deliberately.
Problem & data review
What decision does this software have to support, and what data already exists? Most projects are rescoped here — usually smaller.
Architecture & interfaces
Data model, APIs and screens agreed before code is written, so the expensive decisions are made while they are still cheap.
Iterative build
Working increments you can run yourself, at a fixed cadence. No big-bang delivery at the end.
Evaluation on your data
Accuracy, latency and failure cases measured on your material — not on a public benchmark that flatters the model.
Deploy & hand over
Release, monitoring, source repositories and build instructions. Your team can run and extend it without us.
What you receive
- Source repositories with build and deploy instructions
- Architecture and data-model documentation
- Evaluation report measured on your own data
- Deployment runbook and monitoring setup
- Handover session with your team
Selected work
Work in this discipline
Software
Precision dimensioning from multi-angle images
A Python measurement pipeline that turns five camera views of a part into dimensions with stated uncertainty — and a DIN-style drawing.
Read case study
2026
Software
Multi-tenant ERP for a group of small companies
A Flutter and Supabase business system built for five companies first — and designed to be sold to other SMBs later.
Read case study
2026
Need more than one?
Most projects do. The four disciplines are run by one team, so nothing is lost between a board, its firmware, its housing and the plan that ships it.
Electronics
Custom boards and the software that runs on them — from schematic to a device that boots in the field.
Mechanics
Enclosures, brackets and fixtures designed around the electronics — modelled, printed and fitted, not ordered and hoped for.
Management
Engineering projects steered to a release — scope, owners, gates and honest status, instead of a launch date and hope.
For businesses that need it to work in the field, not in a demo.
Plantations, factories, workshops and infrastructure operators across Indonesia — and product companies in Europe. Tell us the problem; we will tell you honestly whether we are the right team.