Frequently asked questions

What makes us different isn't that we have AI. It's what we bring to AI. Every build is centred on making your business more human: we find where your people are stuck doing robotic work, and bring the intelligence, process, and automation to relieve them of it. The result is people and customers who get to connect, create, and experience. That's what it means to be more human.

Yes. AI isn't the answer to everything. The pros and cons are real. We come in, understand your business, and work out where AI helps you become more human, not less. Sometimes that's an automation, sometimes it's a system your team uses, sometimes it's leaving things alone.

Honestly? You probably are, at least somewhere. Most businesses have at least one workflow where AI would save time or remove friction. The real question is where to start, and what to leave alone. Discovery is how we figure that out: we look at your systems, your team, and where the pressure points are, then recommend the smallest useful first step.

We run a Discovery to understand your business, your team, and where the pressure points are, then we build. That could be a single automation, a custom agent, an internal system, staff training, or a combination. Whatever ships, it's tied to an outcome you can measure.

Both. We're based in Hobart and work with Tasmanian SMBs through to nationally distributed operations. The engagement shape changes with scale; the standards don't.

Our main driver is Claude Opus, with independent review by OpenAI's Codex on production builds. Our C-Suite runs on a deliberately mixed stack: AI Director on GPT, Compliance Officer on Gemini, and local workloads on Google Gemma running on our own hardware. We mix models on purpose: leaning on a single model bakes its assumptions and biases into your business, so our agents cross-check each other across platforms and integrate through our own HQ system.

Yes and no. We enjoy fast, creative building, but 158 Lab codes with structure. Every build ships with the compliance paperwork behind it: ADRs, PIRs, audit logs, documented decisions. You tell us what you need, and you'll have the evidence trail to prove it was built properly.

An automation runs a defined sequence: trigger, steps, done. An agent decides what to do next based on context and the tools it has. We build both, and we'll tell you which one actually fits your problem (often it's the automation).

Yes. 158 Lab runs its own local AI infrastructure on on-premise hardware. For sensitive data like client PII, confidential strategy, or anything restricted, we can keep the whole pipeline in your building or ours. No third-party AI providers touching the data.

Yes. We build agents, voice agents, coded tools, and integrations. We use frontier models, our own in-house local models, or both. From high-level cloud AI to locally-hosted secure AI, we cover the whole spectrum.

Yes. Most of the AI we build connects to something you already use: CRMs, spreadsheets, email, accounting systems, helpdesks, your website. We build on open standards (APIs, webhooks, MCP) so AI slots into your existing stack rather than replacing it.

Yes. Websites aren't our traditional service, but they're often where the AI systems live: chat, lead capture, booking, dynamic content. When a website is part of the AI story, we build it end-to-end on the same stack we use internally.

Yes. We run in-person and online training, but first we go through a Discovery process to understand how your team currently uses AI, what systems you want to apply it to, and how to get there without creating new problems.

Yes. Every build includes a handover, and most clients stay engaged for iteration: fixing what comes up, extending what's working, and keeping systems current as the AI landscape changes. The shape of the ongoing relationship is scoped in Discovery so it matches how you actually work with us.

Depends on scope. A targeted AI automation can land in a week. A custom agent or internal system is usually 4-8 weeks from Discovery to production. Larger transformations run longer and in stages.

That's not how we think about it. AI is the tool, not the advantage. Your people are. We build systems that remove the busywork so your team can focus on what only humans can do.

We classify every piece of data by sensitivity tier and match the AI provider and hosting model to that tier. Confidential business data only goes to providers with proper enterprise privacy controls and no-training commitments. Client PII and sensitive business data stays on approved local infrastructure, not in third-party AI clouds.

Yes. We build with the security discipline you'd expect in regulated environments, informed by frameworks like OWASP ASVS, APRA CPS 234, ISO 27001, NIST SSDF, and DSPANZ SSAM. Not because every client needs full compliance on day one, but because quality, trust, and auditability are how we work.

158 Lab's own core infrastructure sits in Hobart, Tasmania. For client builds we work with you to pick the right hosting model (AU/NZ cloud, your own environment, our infrastructure, or a mix) based on your security, compliance, and data residency needs.

Our core services have from-pricing published on our Services page. For custom AI consulting (agents, integrations, training programs) we scope it in a Discovery session so the price matches the actual work.