Independent engineering software

Precision tools, built one at a time. Desktop and mobile software for design, analysis and the field.

Eleven applications across Windows and Android — CAD and geometry, image processing, strategy backtesting, 3D inspection and GPS navigation. Each one is built for a specific job.

View the work Mobile apps

Projects: 11
Windows: 6
Mobile: 5

01Platform / Desktop

Windows Applications

Desktop tools for authoring, analysis and inspection — built for long sessions, large files and repeatable results.

6 projects

WIN · 01 Released

Aligns and merges a bracket of frames into one image with extended depth of field or dynamic range.
- Python
- PyQt6
- OpenCV
+4
WIN · 01 · scale 1:1

Platform
Windows Desktop

Status

Released

Category
Image Processing

Stack
Python
WIN · 01

Stacker

Focus and exposure stacking

Overview

Stacker takes a burst or bracket of photographs of the same scene and combines them into a single frame. It handles subpixel registration, per-pixel focus measurement and weighted blending, so hand-held sequences and microscope or macro stacks converge without visible seams. Built for photographers and lab users who want repeatable results rather than a one-click preset.
Key features

Subpixel alignment with feature matching
Laplacian and wavelet focus measures
Exposure bracketing and tone mapping
16-bit and RAW pipeline throughout
Batch processing of multiple stacks
Technologies

Python
PyQt6
OpenCV
NumPy
SciPy
scikit-image
exiftool
WIN · 02 In development

Local generative pipeline for upscaling, denoising and restoring photographs on your own GPU.
- Python
- PyQt6
- PyTorch
+4
WIN · 02 · scale 1:1

Platform
Windows Desktop

Status

In development

Category
Image Processing

Stack
Python
WIN · 02

Generative Image Improvement

Restore and upscale

Overview

A desktop front end for running generative restoration models without sending images to a service. Upscaling, denoising, face restoration and detail synthesis chain into a pipeline you can reorder and inspect, with side-by-side comparison at every stage. Model weights, tiling and VRAM budget stay under explicit control.
Key features

Tiled inference for large images
Chainable restoration stages
Side-by-side and split-view comparison
Local model management, no cloud calls
Adjustable VRAM and batch budget
Technologies

Python
PyQt6
PyTorch
ONNX Runtime
CUDA
Pillow
NumPy
WIN · 03 Released

Systematic strategies are tested against history with explicit costs and walk-forward splits.
- Python
- PyQt6
- pandas
+4
WIN · 03 · scale 1:1

Platform
Windows Desktop

Status

Released

Category
Quantitative Finance

Stack
Python
WIN · 03

Backtester and Optimizer

Strategy research bench

Overview

A research environment for testing systematic trading strategies against historical data. Strategies are written in Python and run through an event-driven engine that models commissions, slippage and position sizing explicitly. Parameter sweeps run across grids or Bayesian search with walk-forward splits, and results are reported as a distribution of outcomes rather than a single flattering curve.
Key features

Event-driven execution with cost model
Grid and Bayesian parameter search
Walk-forward and out-of-sample splits
Equity, drawdown and trade analytics
Parallel runs across CPU cores
Technologies

Python
PyQt6
pandas
NumPy
Numba
SQLite
PyQtGraph
WIN · 04 Released

Every part of a glTF or GLB asset, from scene graph to animation tracks, shown in a native window.
- C++
- Qt 6
- OpenGL
+4
WIN · 04 · scale 1:1

Platform
Windows Desktop

Status

Released

Category
3D / CAD

Stack
C++
WIN · 04

GLB Viewer

Inspect glTF assets

Overview

GLB Viewer opens glTF and GLB assets in a native OpenGL window and shows what is actually inside them. Scene graph, meshes, materials, textures, skins and animation tracks are all browsable, with physically based rendering so materials read the way an engine will draw them. It exists to verify an asset before it reaches a pipeline.
Key features

PBR rendering with IBL environments
Scene graph and node inspector
Material and texture channel preview
Animation and skin playback
Wireframe, normals and overdraw views
Technologies

C++
Qt 6
OpenGL
tinygltf
Assimp
GLM
stb_image
WIN · 05 Prototype

Constrained sketches feed a feature tree that stays editable long after the solid is finished.
- C++
- Qt 6
- Open CASCADE
+3
WIN · 05 · scale 1:1

Platform
Windows Desktop

Status

Prototype

Category
3D / CAD

Stack
C++
WIN · 05

CAD Modeller

Parametric solid modelling

Overview

A modelling application built around constrained sketches and a feature tree. Draw a profile, apply dimensional and geometric constraints, then extrude, revolve, fillet or cut, and every operation stays editable afterwards. Models export to STEP and STL for manufacture or printing.
Key features

2D sketcher with geometric constraints
Editable parametric feature history
Extrude, revolve, sweep, fillet, boolean
STEP, IGES and STL export
Python scripting for repeat geometry
Technologies

C++
Qt 6
Open CASCADE
OpenGL
Eigen
Python (scripting)
WIN · 06 Beta

Where rides get measured: GPX and FIT import, elevation and grade profiles, climb and segment stats.
- Python
- PyQt6
- PyQtGraph
+4
WIN · 06 · scale 1:1

Platform
Windows & Android

Status

Beta

Category
GPS & Mapping

Stack
Python
WIN · 06

Bike Routes Viewer

Ride analysis at the desk

Overview

The desktop half of Bike Routes Viewer, where recorded rides and planned routes are imported and pulled apart. Elevation profiles can be scrubbed, climbs are detected and graded, and segments compared between rides on a screen wide enough to hold the whole library. Rides stay in a local library that the Android build reads, so a route worked out here opens the same way on the bike.
Key features

GPX, FIT and TCX import
Scrubbable elevation and grade profile
Automatic climb detection and grading
Segment comparison between rides
Local ride library shared with Android
Technologies

Python
PyQt6
PyQtGraph
GDAL
pyproj
SQLite
GPX / FIT parsing

02Platform / Android

Mobile Apps

Applications that work away from a desk: navigation, mapping and signalling in the field — and one game.

5 projects

MOB · 07 Beta

Pick a runway threshold and the display shows lateral deviation, corridor, ground speed and time to run.
- Kotlin
- Jetpack Compose
- Compose Canvas
+4
MOB · 07 · scale 1:1

Platform
Android

Status

Beta

Category
Navigation

Stack
Kotlin
MOB · 07

GPS Approach

Guidance on final

Overview

GPS Approach turns a phone into an advisory approach display: pick a runway threshold or waypoint and it shows lateral deviation on a course needle, a corridor view of the track that remains, plus ground speed and time to threshold read from the GNSS fix. Heading is fused with the inertial sensors so the needle stays readable between fixes and at low ground speed. It is a situational-awareness aid for VFR and gliding practice, not a certified navigation instrument, and the interface says so plainly.
Key features

Course deviation needle with selectable scale
Runway and waypoint approach corridor
Ground speed and time to threshold
Heading smoothed between GNSS fixes
Readable dimmed and fully offline
Technologies

Kotlin
Jetpack Compose
Compose Canvas
Fused Location Provider
Raw GNSS fixes
SensorManager fusion
WGS84 geodesy
MOB · 08 In development

No engine underneath, just one dragon, parallax night terrain and a game loop written from scratch.
- Kotlin
- OpenGL ES 3.0
- GLSL shaders
+3
MOB · 08 · scale 1:1

Platform
Android

Status

In development

Category
Games

Stack
Kotlin
MOB · 08

Little Dragon

Hand-built side-scroller

Overview

Little Dragon is a side-scrolling arcade game built without a third-party engine: a fixed-timestep simulation loop, a small entity system and OpenGL ES for drawing. The dragon flies through procedurally assembled night terrain with parallax layers, and sprites are batched so the render path stays predictable on older phones. It began as an exercise in knowing what every frame of a mobile game costs, and stayed a game because it was worth playing.
Key features

Fixed-timestep loop, no third-party engine
Procedural parallax night terrain
Batched OpenGL ES sprite rendering
Touch and tilt control schemes
Local progress and best-run storage
Technologies

Kotlin
OpenGL ES 3.0
GLSL shaders
Oboe audio
SurfaceView game loop
DataStore
MOB · 09 Beta

Reading airband frequencies off a map instead of a chart, offline, with nearest-station lookup.
- Kotlin
- Jetpack Compose
- MapLibre GL Native
+4
MOB · 09 · scale 1:1

Platform
Android

Status

Beta

Category
Aviation Radio

Stack
Kotlin
MOB · 09

AirBand Map

Airband frequencies on a map

Overview

AirBand Map puts VHF airband frequencies where they are easiest to read, on the map itself. Tower, ground, approach, ATIS and en-route sector channels are tied to the airfield or airspace they belong to, so finding the right frequency is a glance rather than a chart lookup. Frequency and airspace data is imported into a local database and tiles are cached by region, so it works with the radio off and the network gone.
Key features

Airband frequencies on an offline map
Tower, ground, approach and ATIS channels
Airspace and sector boundary overlays
Nearest-station lookup from current position
Optional live aircraft track layer
Technologies

Kotlin
Jetpack Compose
MapLibre GL Native
MBTiles offline tiles
Room
Kotlin Coroutines
Retrofit
MOB · 10 Released

When the phone is the only light to hand, it signals: morse, SOS and timed torch patterns.
- Kotlin
- Jetpack Compose
- Camera2 torch control
+3
MOB · 10 · scale 1:1

Platform
Android

Status

Released

Category
Outdoor & Safety

Stack
Kotlin
MOB · 10

Night Signal

A phone as a beacon

Overview

Night Signal drives the camera torch and the screen as a controllable light beacon: morse encoded from typed text, SOS, slow strobes and custom on/off groups. A foreground service keeps a pattern running with the screen dark, so the phone can be propped up and left signalling. Timing, duty cycle and brightness are exposed directly rather than hidden behind presets, because the decision that matters is how much battery a given signal is worth.
Key features

Morse encoding from typed text
Torch and full-screen strobe output
Adjustable timing and duty cycle
Keeps signalling with the screen off
Preset SOS and attention patterns
Technologies

Kotlin
Jetpack Compose
Camera2 torch control
Foreground Service and WakeLock
Kotlin Coroutines
DataStore
MOB · 11 Beta

On the handlebars, the same routes run offline with an elevation profile and a live position cursor.
- Kotlin
- Jetpack Compose
- MapLibre GL Native
+3
MOB · 11 · scale 1:1

Platform
Windows & Android

Status

Beta

Category
GPS & Mapping

Stack
Kotlin
MOB · 11

Bike Routes Viewer

Offline routes in the field

Overview

The Android half of Bike Routes Viewer, made for the ride rather than the desk. It draws the route over offline tiles with an elevation profile, gradient bands and a cursor that ties current position on the track to the climb ahead, with controls sized for gloves and glare. Route parsing and geometry are shared with the Windows build, and tiles and rides live on the device, which is the whole point once the signal goes.
Key features

Route and profile over offline tiles
Gradient bands along the track
Live position cursor on the profile
Rides shared with the Windows build
Whole library stored on the device
Technologies

Kotlin
Jetpack Compose
MapLibre GL Native
MBTiles offline tiles
Room
GPX / FIT parsing
Next build · in design

Precision tools, built one at a time. Desktop and mobile software for design, analysis and the field.

Stacker

Stacker

Overview

Key features

Technologies

Generative Image Improvement

Generative Image Improvement

Overview

Key features

Technologies

Backtester and Optimizer

Backtester and Optimizer

Overview

Key features

Technologies

GLB Viewer

GLB Viewer

Overview

Key features

Technologies

CAD Modeller

CAD Modeller

Overview

Key features

Technologies

Bike Routes Viewer

Bike Routes Viewer

Overview

Key features

Technologies

GPS Approach

GPS Approach

Overview

Key features

Technologies

Little Dragon

Little Dragon

Overview

Key features

Technologies

AirBand Map

AirBand Map

Overview

Key features

Technologies

Night Signal

Night Signal

Overview

Key features

Technologies

Bike Routes Viewer

Bike Routes Viewer

Overview

Key features

Technologies