Why AcesForge Is Two Divisions: Software Consulting and Hardware Evidence Engineering

One Engineer, Two Disciplines, One Standard

AcesForge looks, at first glance, like two unrelated businesses sharing a name. Division B builds software: AI-augmented systems architecture, integration engineering, custom applications. Division A does something that sounds nothing like software at all: bench testing electronics, characterising prototypes, and producing signed evidence reports for patent attorneys and investors. Most consultancies pick one lane and stay in it. AcesForge runs both, deliberately, because they are built on the same underlying discipline: reliability engineering, applied to two different kinds of systems.

That discipline did not start at AcesForge. It started at Tait Communications, where founder Toby Bell-Ramsay has spent four and a half years and counting building reliability-critical radio communications systems, the kind where a comms failure during an emergency response is not an acceptable outcome. That background earned the Steven Bly Award for Most Outstanding Product or Solution Development in 2025. AcesForge applies the same standard, instrumented, documented, verified, to client work in software and hardware alike.

Division B: Software and Systems Architecture

Division B is AI-augmented software and systems architecture: designing the structure of a system before writing a line of it, then building fast with AI leverage while keeping a human-verified reliability boundary around every AI-assisted step. The entry point is a $1,500 fixed-price Systems Architecture Review: a written architecture map covering what the system should be, where the integration seams and failure modes sit, where AI automation is safe to trust, and where a human has to verify. The review ends with a named, priced first build phase, so there is a concrete next step before the engagement even ends. Build work past that point runs at $220 to $280 an hour, committed-hours, not open-ended.

This division exists because most software money gets wasted on the build before anyone has properly designed the system. AcesForge designs first. The AI-augmented part is not a marketing label: it means one engineer holding more breadth and moving faster than a same-cost human-only team would, inside a stack of TypeScript, Python, React and Next.js, Cloud Run, and Firestore, with every AI-assisted capability paired to an explicit reliability boundary rather than trusted by default.

Division A: Hardware R&D and Evidence Engineering

Division A is independent hardware R&D and evidence engineering: pre-compliance bench testing, prototype characterisation, and signed test reports for NZ inventors and electronics firms who need evidence that stands up to a patent examiner, an investor, or a certifier. AcesForge is not an accredited EMC lab and does not issue accredited certification reports. What it does is the design-stage work that happens before that: bench-level emissions and susceptibility checks, device behaviour mapped across voltage, temperature, load, or frequency ranges, and reliability and stress testing that shows how a device performs across conditions, not just at nominal.

The engagement shape is fixed-scope and phased, starting at $750 for an Evidence Gap Audit that maps a specific product against the mandatory AS/NZS and CISPR standards that apply to it, and identifies exactly what evidence already exists and what is still missing. From there, bench testing and characterisation phases are scoped and priced individually, typically landing between $750 and $9,000 per phase depending on device complexity. Every phase ends with a signed report and a raw data package, methodology, conditions, results, and the measurement files behind them, so any engineer can independently verify the work.

For an early-stage hardware team, the alternative is hiring a salaried test engineer, which runs roughly $80,000 to $100,000 a year in NZ before equipment, bench space, and setup time. For episodic test needs, contracting the evidence production out is usually the more sensible call: AcesForge can start before a company has filled that role internally, and hand over a clean evidence pack once it does.

Why the Same Person Runs Both

The unifying thread is not the subject matter, it is the method. Both divisions apply the same instrumented, documented, verified approach that reliability-critical radio systems demand: define what correct behaviour looks like, measure against it, record the evidence, and make sure someone other than the person who built it can check the work. A software architecture review states explicitly where AI is trusted and where a human verifies. A bench test report states explicitly what was measured, under what conditions, with what result. Same discipline, different systems.

How to Start With Either Division

Division B starts with a $1,500 fixed-price Systems Architecture Review. Division A starts with a $750 Evidence Gap Audit for hardware, or a scoped bench-testing engagement if the gap is already known. Both are Christchurch-based and available for wider Canterbury and NZ engagements, and both end with a named, priced next step rather than an open-ended proposal.