We build production AI systems for teams across Aotearoa — agents that carry real work, and assistants grounded in your own material that answer with citations instead of guesses.
Plenty of AI projects look great in a demo and stall on the way to production. We build for what comes after launch: systems that answer from your data, cite their sources, and keep doing the actual work — triage, briefings, reviews, operations.
Sound familiar?
The problems we're asked to solve
Nobody wakes up wanting AI. They wake up with one of these.
“We waste hours every week on admin, email and double-handling.”
Workflow automation and agents that carry the repetitive work — connected to the tools you already use, with a human in the loop where it matters.
“Customers ask the same questions at all hours.”
A grounded assistant on your site, trained on your own content — it answers with citations, captures the lead, and hands off to a human when it should.
“Our team can't find anything — the knowledge is buried in documents.”
An internal knowledge base your staff can ask in plain English, built on your policies, procedures and project history.
“We know AI matters, but we don't know where to start.”
A fixed-price AI Readiness Audit: where AI genuinely helps your business, what it costs, what to build first — and what to skip.
“We don't want our data going offshore.”
A data-handling summary with every build — where every byte goes, documented and disclosed before anything ships.
“We can't afford an enterprise AI project.”
Fixed prices and staged builds. Start with the smallest thing that pays for itself, then grow it.
What we build
What we're asked for most
/01AI agents that plan, use tools and carry a task to completion
/02Business workflow automation — your existing tools, connected and carried
/03LLM & generative-AI integrations into your product or workflow
/04Retrieval-grounded (RAG) assistants that cite their sources
/06The honest advice layer — including when you don't need AI at all
Where to start
Fixed price, fixed scope, no hourly billing
You’ll know what you’re buying. Every engagement starts with a free 30-minute discovery call — and a price agreed before any work begins.
Delivered in under a week
AI Readiness Audit
A structured discovery session and a written report: where AI genuinely helps your business, realistic return on the investment, and a prioritised roadmap — yours to act on with us or anyone else.
Live in one to two weeks
Grounded Website Assistant
An assistant trained on your own content that answers customer questions around the clock — with cited sources, lead capture, and a clean handoff to a human when it should stop talking.
Running in two to three weeks
AI Content Pipeline
A system that drafts, schedules and publishes content in your brand voice — connected to your CMS and channels, with your review built into the loop.
Custom builds — internal knowledge bases, workflow automation, document processing, agentic systems — are scoped per project, with a fixed price and clear scope before we start.
Who we build AI for
We build the same way for a two-person startup and a health network.
Startups & SMEs
AI that ships and holds up in production — not a proof-of-concept that stalls. We’ll tell you the smallest thing worth building first.
SaaS & product teams
AI features added to an existing product without destabilising it — integrated into your codebase, your data, your release process.
Government & public sector
Grounded, cited, safety-tested systems for high-trust environments — like the AI we built inside a New Zealand health network.
Agencies & teams
An embedded AI partner behind your brand when the work needs capability you don’t have in-house. More on partnering →
AI you can stand behind
Every build ships with its paperwork
Every AI system we deliver ships with its governance built in. If you ever have to explain your AI to a customer, a board, or a regulator — you’ll have the answers in writing.
We work to the Privacy Act 2020, MBIE’s Responsible AI Guidance, and — for public-sector work — the Public Service AI Framework. Every project includes:
/01A documented purpose statement — what the AI does, and doesn't
/02A privacy impact assessment for personal information
/03A data-handling summary — where your data goes, disclosed
/04Transparency — people know when they're talking to AI
/05A human oversight plan — who checks, and when
/06Bias and safety review before go-live
Proof
From work that shipped
We ship what we use. Ask Pinnacle answers health questions inside a large New Zealand health network — red-teamed against 80-plus scenarios before launch. Morris, our own AI operations agent, helps run this studio.
Production AI systems — AI agents that do real work, LLM and generative-AI integrations, retrieval-grounded (RAG) assistants, and the governance and responsible-deployment layer around them. The kind of AI that has a real job to do and has to keep doing it after launch.
Production AI systems — AI agents that do real work, LLM and generative-AI integrations, retrieval-grounded (RAG) assistants, and the governance and responsible-deployment layer around them. The kind of AI that has a real job to do and has to keep doing it after launch.
Both. We build for startups and SMEs who want AI that actually ships, for SaaS and product teams adding AI to an existing product, for government and public-sector projects, and as an embedded AI partner inside other agencies. The engineering discipline is the same at every scale.
Both. We build for startups and SMEs who want AI that actually ships, for SaaS and product teams adding AI to an existing product, for government and public-sector projects, and as an embedded AI partner inside other agencies. The engineering discipline is the same at every scale.
It depends on what you're building — a readiness audit is a very different engagement to an agentic system. What's always true: we scope every project up front and quote a fixed price with a clear scope before any work starts. No hourly billing, no surprise invoices. A free 30-minute discovery call is the fastest way to get a real number for your situation.
It depends on what you're building — a readiness audit is a very different engagement to an agentic system. What's always true: we scope every project up front and quote a fixed price with a clear scope before any work starts. No hourly billing, no surprise invoices. A free 30-minute discovery call is the fastest way to get a real number for your situation.
With a free 30-minute discovery call. Tell us what you're stuck on and we'll say honestly whether AI helps, what it would cost, and what to build first. If you want it in writing, the AI Readiness Audit turns that into a prioritised roadmap you can act on with us or anyone else.
With a free 30-minute discovery call. Tell us what you're stuck on and we'll say honestly whether AI helps, what it would cost, and what to build first. If you want it in writing, the AI Readiness Audit turns that into a prioritised roadmap you can act on with us or anyone else.
Wellington, Aotearoa New Zealand — in your timezone, reachable in your working day, no offshore handoff.
Wellington, Aotearoa New Zealand — in your timezone, reachable in your working day, no offshore handoff.
We red-team what we build before anyone else can. Ask Pinnacle, an assistant we built for a New Zealand health network, faced 80-plus attack and safety scenarios before it went live. Grounded answers, cited sources, and an honest 'I don't know' over a confident guess.
We red-team what we build before anyone else can. Ask Pinnacle, an assistant we built for a New Zealand health network, faced 80-plus attack and safety scenarios before it went live. Grounded answers, cited sources, and an honest 'I don't know' over a confident guess.
Every project ships with a data-handling summary that documents exactly where your data goes — including any offshore APIs a model calls — before anything ships. Nothing leaves the country without being disclosed and signed off. Where data residency matters, we design for it, in line with the Privacy Act 2020 and its cross-border disclosure rules.
Every project ships with a data-handling summary that documents exactly where your data goes — including any offshore APIs a model calls — before anything ships. Nothing leaves the country without being disclosed and signed off. Where data residency matters, we design for it, in line with the Privacy Act 2020 and its cross-border disclosure rules.
Yes. We build grounded, cited, safety-tested systems for high-trust environments — the kind you're audited against. We work to the standards the New Zealand Public Service AI Framework and the Government's Responsible AI guidance set out, and we've delivered inside a New Zealand health network alongside public-sector organisations.
Yes. We build grounded, cited, safety-tested systems for high-trust environments — the kind you're audited against. We work to the standards the New Zealand Public Service AI Framework and the Government's Responsible AI guidance set out, and we've delivered inside a New Zealand health network alongside public-sector organisations.
A chatbot answers questions. An agent does work — it takes a goal, plans the steps, uses tools, and carries a task to completion with a human in the loop where it matters. Right now, a lot of the real value in AI is in agents rather than chat windows.
A chatbot answers questions. An agent does work — it takes a goal, plans the steps, uses tools, and carries a task to completion with a human in the loop where it matters. Right now, a lot of the real value in AI is in agents rather than chat windows.
Start with a conversation
Free 30-minute discovery call
Thinking about AI? Book a free 30-minute discovery call. Describe what you’re stuck on and we’ll come back with how we’d approach it, what it would cost, and whether AI is even the right tool.
