Quantum-optimized propulsion hardware
Mission profile in. Machined module out.
8 → 2
Months to production (industry vs. PROMETHEUS)
3
Active design partners on the pipeline
The problem

Propulsion engineering runs on iteration. We replaced iteration with optimization.

Legacy primes move through design, prototype, test, scrap, repeat. Each loop costs months and millions. A single engine module can take the better part of a year to reach production.

01
Classical solvers cannot explore the full design space. They hit local minima. Engineers fall back on tribal knowledge and overbuilt safety factors.
02
CAD-to-manufacturing is a manual translation. Geometry is reinterpreted, toolpaths are hand-written, errors are introduced at every seam.
03
No feedback loop. Dimensional inspection data from the machined part never reaches the simulation layer. The next design starts from zero.
Industry standard
8–12 months
Per engine module. Multiple design-test-scrap cycles. Dedicated CAM programmers. High material waste.
PROMETHEUS
60 days
Billet to inspection-ready module. Single pass. Zero prototype scrap. Closed-loop inspection data feeds the next design.
Pipeline

Four steps. No handoffs. No translation errors.

Every stage is proprietary. Every stage was built by the founding team. No third-party CAD, no external CAM vendors, no outsourcing.

Step 01

Mission intake

Customer submits thrust profile, envelope, and thermal constraints. Our parameterization engine maps the full design space in seconds.

Step 02

Quantum solve

Nozzle contour and topology are formulated as QUBO problems and solved on D-Wave QAOA. Alloy chemistry runs on IBM VQE. Free simulators by default; QPU on demand.

Step 03

Toolpath generation

Our PicoGK-based voxel engine converts the quantum-optimized geometry directly into 5-axis G-code. No CAD translation. No human CAM.

Step 04

Machine & deliver

In-house 5-axis CNC cuts Inconel, Titanium, or Aluminum to spec. Every dimension is inspected. Data loops back into the quantum model for the next run.

Technology

Four layers, one stack, zero middlemen

Everything from the quantum optimizer to the CAM toolpath generator was written in-house. No black boxes, no licensing dependencies.

Quantum optimization layer

QAOA & VQE solvers

Python sidecar connecting to D-Wave Leap and IBM Quantum. Nozzle contour, topology, and alloy chemistry are each formulated as the native problem type — QUBO for combinatorial, variational for molecular.

D-Wave Advantage · IBM Heron · Qiskit · Ocean SDK
Geometry & physics kernel

Voxel-computational engine

.NET-based PicoGK kernel with custom propulsion extensions: thermal field mapping, rotational stress, manufacturability constraints. Everything in voxel space — no mesh, no CAD artifacts.

PicoGK · Custom FEA solvers · Signed distance fields
Manufacturing interface

5-axis CAM toolpath generator

Consumes quantum-optimized voxel geometry directly. Generates roughing, semi-finish, and finish passes optimized for Inconel 718, Ti-6Al-4V, and 6061-T6.

Custom G-code · In-house 5-axis CNC · In-process probing
Desktop application

Tauri + Rust + Three.js

Cross-platform desktop app for monitoring quantum runs, inspecting voxel geometry slice-by-slice, and reviewing manufacturing reports before the toolpath is cut.

Rust · WebGPU · Real-time voxel rendering
Validation
60days
From raw billets to a fully machined GE T700 cold section module

We picked the GE T700 because it is one of the most complex turboshaft assemblies in service — tight tolerances, thin walls, complex internal flow passages. We machined it start-to-finish using our own quantum-optimized pipeline. No legacy CAD. No manual CAM. No scrapped prototypes. Every dimension verified against the OEM manual.

3
Active design partners running mission profiles through PROMETHEUS today
2
Signed LOIs for production purchase orders upon STL validation
~100%
First-pass machinability rate — zero wasted stock, zero rework
$18k
Per-unit contract value for a custom impeller and combustor assembly
Founders

Two operators. Equal equity. Six years of building together.

We met on a turbomachinery R&D project. We have worked side-by-side on the shop floor. Both full-time, both all-in, since day one.

Co-founder & CEO

Balaraju M

Core Rust/Tauri frontend · Quantum optimization · Proprietary geometry solver

Built the algorithmic core from scratch — the quantum-to-CAD pipeline, the voxel geometry kernel, and the desktop application. Full-time since the project started.

Co-founder & CTO / Head of Manufacturing

Prajwal B G

Thermodynamics physics models · CAM toolpaths · 5-axis CNC operation

Former aerospace engineer. Quit his role to run the shop floor full-time. Writes the physics simulation layer and operates every machine. His domain expertise is the difference between a part that looks right and a part that survives 20,000 RPM.

Design partners

Three mission profiles running today

Each partner has submitted real propulsion requirements. We are currently in the mission-to-STL validation phase. The first paid unit is scheduled to ship in 8–10 weeks.

Defense startup — drone engines
Defense startup — small launch vehicles
University propulsion research lab
Infrastructure

In-house stack. Industrial-grade.

Every capability from quantum compute to 5-axis chip formation lives under one roof. No subcontractors, no lead times, no IP leakage.

Quantum backends
D-Wave Advantage QAOA
IBM Heron VQE
Qiskit + Ocean SDK Hybrid
Manufacturing
5-axis CNC ±5 μm
4th-axis rotary In procurement
In-process probing Closed-loop
Materials
Inconel 718 Approved
Ti-6Al-4V Approved
6061-T6 Aluminum Approved
Contact

We build quantum-optimized propulsion hardware for defense and aerospace.