Independent Development Studio
We build AI platforms, industrial field tools, mobile apps and games — from a food-rescue network serving North America to a multi-provider AI desktop suite that runs entirely on your own machine. Nine products across four platforms — three live, the rest in beta or active development.
The Quanta Foundational Code of Ethics
Spoken in Quanta's own voice — the same one the product speaks with. The geometry is a true Flower of Life: nineteen circles, each struck on an intersection of the ones before it.
Installed, not illustrated
Every frame below is a capture of the app actually running — no renders, no mockups, no store art. Switch between them and step through the screens.
Measured, not claimed
Every figure below is counted directly from the source tree — excluding dependencies,
build output and generated files. Re-run npm run stats to regenerate it.
The portfolio
Everything we've built — tagged by where it actually stands. Nothing here is a mockup or a landing page waiting on a product. Production means live and in users' hands. Beta means installable and working, still being hardened. In development means it runs, with work remaining.
A completely independent, multi-tier AI model suite running directly on private clusters and local edge devices — zero external telemetry, zero corporate lineage, and transparent step-by-step reasoning. Governed by the immutable Quanta Foundational Code of Ethics.
Every other app tells you what is there; this one tells you whether you can live there. Sweep the phone across the horizon and the camera builds a measured obstruction map — 72 bearings, each carrying the elevation above which the sky is clear. Crossed with a from-scratch NOAA solar model it reports the watt-hours a site will actually deliver, not the brochure figure: a 45° treeline costs a 600 W array around 40% of its day, and a treed site feels fine right up until the battery doesn't last the week. New in 2.3, every pin is coloured by whether your rig holds four nights there, and the corridor planner runs the same arithmetic over a whole route, night by night. Starlink passes come from SGP4 on live CelesTrak elements. 32,685 campsites, 308,102 designated routes and 652,600 public-land polygons — legal access modelled per vehicle class and per jurisdiction, and unknown wherever the source declines to say. And now it drives you there: offline turn-by-turn navigation on downloadable per-state graph packs, with the computed line read back against the MVUM and BLM designations it crosses. It works with the radio off, and where nothing has been measured it tells you so rather than filling the gap.
A vertical climb through generated dimensions, with synchronised multiplayer sectors that show other pilots' altitude live on your HUD. Describe a universe in plain language and the rift engine builds it — heavy gravity, neon platforms, whatever you ask for. There's an in-game Oracle you can genuinely consult on run strategy, a prestige ladder that trades progress for permanent physics multipliers, and a tiered emporium of trails and avatars.
An AR food scanner that refuses to guess. Point the camera and a luminance-and-variance guard throws out a dark or obstructed frame before any model runs — the same “measured or nothing” rule the rest of this studio is built on. What passes goes to Gemini vision and a 3M-product Open Food Facts lookup for real detection. From there it opens out: a metabolic digital twin that simulates the 120-minute glucose curve, a toxin X-ray across 40+ additives and seed oils, food-pairing for absorption (vitamin C lifts iron uptake, piperine lifts curcumin), a price-arbitrage router that splits one shop across eight chains for the exact saving, an AR fridge microclimate map, and a neighbourhood barter mesh. All on-device, offline-first.
A food- and supply-rescue network that connects surplus to the people who need it, across North America. Agentic matching pairs donations with shelters, food banks and crisis services in real time. Ships as a web app, an installable PWA, and native iOS and Android builds off one codebase — with its API running on Cloudflare Workers at the edge.
Five AI providers in one Windows app — Gemini, Claude, OpenAI, and local Ollama or LM Studio servers, each with live latency on a provider pulse. It calls those providers directly: no server of ours in between, no account, nothing synced. Keys are sealed with Windows DPAPI and the interface only ever sees the last four characters. Grounding tools, a local IDE with a coding agent, an executing workflow graph, audit logs and snapshots that survive restarts.
Flat-pattern layout for industrial insulators — gore elbows, tees and laterals developed to the ordinate, then proved: the pattern rolls up in 3D and only closes if the ordinates are right. A 2D sheet-metal nesting engine bin-packs the blanks and reports the scrap yield; a spool-assembly builder stacks multi-fitting runs into one take-off; and an ASTM C680 pass sizes insulation against the dew point to keep condensation and CUI off the steel. Field material and fastener counts, R-value and cost fall out of the same model. Built for Alberta SAGD steam distribution and Canadian Red Seal compliance.
A social client with no engagement target, and the architecture to back the claim. Posts, likes and the feed are held in Room on the device; the same store renders as a tiled grid or a single-column stream. An Attention tab counts session time on the phone and stores it nowhere, and its controls only ever subtract. Integrations that have no credentials render an Unconfigured state rather than plausible-looking output — an idea worth stealing. Demo and release are one codebase behind a compile-time flag, so the demo branches are not present in a release build at all.
A full DAW built around generative composition — a Lyria-driven track builder, multi-track arranger with an AI section editor, six-knob macro rack, mixer console with per-channel sidechain and VST slots, spectral analysis and a master bus. Exports audio, video and MIDI.
Turns the coding agent you're already talking to into an orchestrator. One tool call puts a task to five specialists at once — architect, optimizer, sentinel, builder, executor — each backed by its own local Ollama model. They aren't one-shot prompts: every member gets read, list and search tools and investigates the workspace itself, so reports cite real code. They run concurrently, so consulting five costs about what consulting one does, and only the orchestrator may write — a member reaching for a write tool is refused, which is why the whole team also works in plan mode.
Real screens, real builds
No mockups — the screens below are captures from the running applications on a working machine, next to the launch films we cut for the ones now shipping.
Set genre, mood, key, vocal style and groove, then let a Lyria model build the track — and take over the moment you want to. Everything after generation is a real DAW.
Gemini, Claude and OpenAI when you want them — Ollama and LM Studio when you don't. Requests go straight from your machine to whichever you pick, with no server of ours in between, no account and nothing synced. Run it entirely on local models and nothing leaves the box.
Insulation work happens on pipe racks in the cold, in gloves, one-handed. That's the constraint the whole product is designed around — and the reason the maths has to be right the first time.
Every frame of this is screen capture from a real run — the climb, the Oracle, the emporium, the live multiplayer rifts. The captions are typeset over it; nothing on screen was generated, redrawn or re-lettered.
com.cosmicascension.myapp, v1.4.0The launch teaser for Quantity — a produced title piece, not a screen capture. The app itself is the six screens in its card above; this is the thirty-second version of what it is for.
The launch film for Roam & Respawn — a produced feature tour that closes on the real app map. It sells the one thing no competitor can copy: the sky measured over your head, crossed with the sun and the satellites.
com.aistudio.roamrespawn.nomadThe speed control
Most AI tools give you one gear, and it's the slow one. Gemini's thinking models deliberate before they answer — brilliant on a hard problem, agonising when you already know what you want. QuantaSync puts that on a slider. Drag it to the last stop and the model stops deliberating entirely: it answers the instant it can.
This panel is a recreation of the app's behaviour, timed from measured runs — the page makes
no model calls of its own. Inside QuantaSync the slider sets Gemini's
thinkingLevel, and the
last stop sends MINIMAL
rather than omitting the config — on a thinking model, leaving it out means "model default",
which is not the same as asking it to stop.
Live in your browser
Five demos across four of our products, all running on your machine. Nothing here is a video and nothing here is faked: the flat pattern is real trigonometry, the heat loss is a real iterative solve, the sky plot is real orbital geometry and the spectrum is your own audio through a real FFT.
| Satellite | Bearing | Elev. | Range | Delay | State |
|---|
Straight about what this is: the five
loops are pre-rendered. They were synthesised offline by scripts/render-tracks.mjs
— oscillators, filtered noise and a sidechain envelope, no samples anywhere — and shipped as
files. Pressing Generate picks one you have not heard and plays it. The staged progress is a
reconstruction of what Quantalize does on the desktop, where a Lyria model actually composes
to the brief. Nothing is generated in this tab.
The analysis, however, is entirely real.
Whatever plays — a generated loop, your microphone, or a file you load — goes through a
4096-point FFT, band energies summed over the bins those frequencies actually occupy, and a
tempo recovered by autocorrelating a 50 Hz onset envelope. The recovered BPM should land
on the tempo printed beside the track title, which is a check you can make yourself. Nothing
is recorded and nothing is uploaded. One generation per visitor per day, enforced at the edge
against a salted, self-expiring hash — never a stored address.
State-of-the-Art Architecture
Five browser capabilities, each running against a real Web API rather than a mock: WebGPU tensor compute, a language model resident on your own GPU, a 3D spatial visualiser, measured edge latency, and the Web Speech API. Open DevTools while you use them — that's rather the point.
A real WGSL compute shader dispatched to your graphics adapter, running the single-precision matrix multiply that every neural network layer is built from. The same multiply is then recomputed on the CPU and the two results are compared element by element — so the speedup on screen is measured on your hardware, not quoted from ours.
// Pick a matrix size and press Run. The shader below is dispatched to your GPU,
// then the same multiply is recomputed on the CPU and the two are compared
// element by element. Nothing here is pre-written output.The card above runs the primitive. This runs the whole stack: a real transformer, loaded into your GPU and executed here. Weights download once from Hugging Face and are cached by the browser — after that it works with the network off. Open the Network tab while you chat and you will see no requests, because there is no server to make them to.
Nothing downloads until you press Load. The model is a few hundred megabytes — small enough to be quick, small enough that it will sometimes be wrong. That is the honest trade at this size, and it is running on your GPU either way.
The portfolio as a graph, projected and rotated in software — no 3D library, just a perspective divide and a depth sort. Drag to orbit; click a node to read what that product actually is.
Live edge endpoint latency, browser heap monitoring, and hardware thread concurrency stats.
Control website navigation and hear studio status updates using real-time voice commands.
"Go to work" "Run local ai" "Telemetry" "Contact"
What we do
Ten things this studio actually does, each tied to something shipped. Every claim below names the product it came out of, so you can go and check it.
QuantaSync talks to Gemini, Claude, OpenAI, Ollama and LM Studio from one desktop app, direct from the machine with no server in between. Keys are sealed in a Windows DPAPI vault and the interface only ever sees the last four characters. Latency is measured continuously across all five.
Hive Mind fans one task to five specialists at once, each on its own local Ollama model with read, list and search tools, so reports cite real code rather than guesses. Only the orchestrator may write — a member reaching for a write tool is refused, which is why the whole team also works in plan mode.
Kotlin and Compose, with the geometry done properly: Roam & Respawn's Sky Scan solves camera bearing and elevation out of the device rotation matrix to build a measured obstruction map. MapLibre tiles, Room caches that survive no signal, CameraX, and API budgets that respect a quota.
Glean ships as a web app, an installable PWA and native iOS and Android builds off a single React and Capacitor source, with its API on Cloudflare Workers at the edge. Shared logic stays shared; only the delivery shell changes.
Cloudflare Workers, Pages, D1 and Email at the edge; Firebase and Google Cloud where they fit. Secrets stay server-side and keys never reach the bundle — including on this page, whose contact form and interest survey are both Pages Functions.
Quantalize is a full DAW: multi-track arranger, six-knob macro rack with MIDI learn, mixer with per-channel sidechain and VST slots. NeuraSync runs genuine FFT on the master bus, band energy from real bins, and tempo from onset-interval autocorrelation — and states plainly where the science stops.
InsuLayout Pro develops lobster-back elbows to the ordinate and solves the ASTM C680 heat balance by iteration, because the surface coefficient and the jacket temperature each depend on the other. Wrong by an inch is a re-cut; wrong by a degree is condensation under cladding.
Cosmic Ascension generates its dimensions from a plain-language description and synchronises sectors so other pilots' altitude appears live on your HUD. On the web: WebGL shaders, WebGPU compute, and the raymarched hypersurface at the top of this page.
When something can run locally, it runs locally. QuantaSync keeps keys, conversations and logs on your own disk. Sovereignty Suite blocks trackers without reading your history. The language model on this page downloads to your GPU and is released again after ninety seconds idle.
Token counts, spend and latency per model accumulated from real requests and exportable to CSV. The statistics on this page are counted from the source tree by a script, never typed. If a number appears anywhere in this studio's work, something measured it.
The studio
QuantaCore Labs is an independent development studio. Small, fast, and unusually unwilling to ship something that doesn't work.
Every product on this page exists as running software with an application ID, an installer or a store package. We label R&D as R&D. Nothing is announced before it runs.
QuantaSync keeps keys, conversations and logs on your own disk and talks to providers directly — no account, no server of ours in the middle. Sovereignty Suite blocks trackers without reading your history. When something can run locally, it runs locally.
Camera field-of-view geometry, lobster-back development, non-blocking multi-model orchestration. We take the problems where getting it approximately right isn't good enough.
A food-rescue network for shelters. A layout tool for insulators on a steam line. Software aimed at real work, not at a demo reel.
Support the studio
Right now this entire studio operates out of a humble basement setup — crafting every algorithm, tuning every coordinate matrix, and designing every screen by the vision of my dream.
The ultimate goal is simple: take this entire development studio fully off-grid and on the road. Living the nomadic reality, field-testing every tool in the deepest wilderness, and proving that one dedicated builder can push meaningful technology to the world from anywhere.
We don’t build software to extract attention. We build to give people autonomy, dignity, and the freedom to thrive.
If any of these tools have helped you, if you believe in technology built with authentic purpose and heart, or if you just want to fuel the next late-night coding session — grabbing a coffee means the world.
Thank you for standing behind independent creators, and for believing in what’s possible when technology is aligned with love, truth and human elevation. Let’s build the future together.
Get in touch
Whether it's a mobile app, a desktop platform, an AI system that has to run on your own hardware, or a field tool nobody has built yet — tell us what you're trying to make.