About
Two people. Both senior. Both on your project.
Silky Insights is a New Zealand applied-AI consultancy and product studio, founded in 2021, delivering for clients across New Zealand, Australia and the United Kingdom. Two people do the work.
There is no bench, no juniors learning on your budget, and no handover gap between the people who advise and the people who build. That is a deliberate constraint, not a stage we are trying to grow out of.
Who we are
Chris Matthews
Co-founder · Strategy, Product, Commercial
chris.matthews@silkyinsights.com+64 21 432 321Chris trained as a mechanical engineer before moving into AI, and the habit stuck: take the thing apart, find where it actually breaks, fix that rather than the symptom. He has spent over eight years on client-facing AI discovery, and most of that work happens before any code is written. What is worth building, whether the data can support it, and what order the work should happen in.
He has run AI strategy, adoption and delivery end-to-end at board level in a regulated New Zealand organisation, and he runs the discovery workshops that help CEOs and boards work out what is feasible, what to build versus buy, and what they can adopt responsibly. Through Silky Insights he has delivered semantic search and RAG platforms, workflow automation and LLM-enabled decision-support products.
Working in regulated environments taught him to build governance in from the start rather than bolt it on at the end, usually alongside external regulatory and privacy stakeholders. Every system he has shipped carries evaluation loops, escalation rules and human oversight. He has also spent five years building Silky Insights from a standing start into a profitable practice, which teaches you fairly quickly what is worth doing and what is not.
Luke Doyle
Co-founder · Chief Technology Officer
Luke builds the systems, and then keeps them running. His work covers the full path from unstructured source material to something a product can query: document ingestion, extraction from PDFs and tables, foundational database design, vector search, and the retrieval architecture that decides what a model actually sees.
He specialises in RAG systems and structured outputs: getting reliable, traceable results out of models rather than plausible-sounding ones, which is what makes AI usable in compliance and other high-stakes settings. He started by building a chess-playing bot with machine learning written from scratch, which is a fair summary of how he still works, down to the mechanism rather than the abstraction.
He has been the technical lead on every Silky engagement: Enerlytica’s data foundations and semantic search, the retrieval and structured outputs inside Humaneer’s Mak, BusinessDesk’s archive search, and the collection layer behind the FOS competitor briefs. The earliest of those has been in production since 2023 and is still supported today. He deploys into infrastructure the client controls across Azure, GCP and AWS, owns the evaluation and monitoring that tells you whether retrieval is genuinely working, and explains what the system is doing in plain language to whoever needs to understand it.
- API calls served
- 20M+
API calls served
Estimated run-rate since October 2023, as at August 2026.
- Documents indexed
- 100,000+
Documents indexed
Indexed across our search systems, as at August 2026.
- Longest system in production
- 4+ years
Longest system in production
Still live, still supported.
- Uptime SLA
- 99.9%
Uptime SLA
Our committed service level.
The pattern
What the work looks like, every time
Take messy, high-value information (report archives, PDFs, data tables, transcripts, policies, recurring data flows), turn it into an AI-ready structured layer, then wrap that layer in a workflow-specific tool: semantic search, question answering, recommendations, portals, reporting pipelines, compliance assistants. Then support it as it becomes operationally central.
Where we work
Regulated environments, where being approximately right is not enough
We work in finance, healthcare and energy, alongside media, HR and compliance. The common thread is regulation and consequence: advice has to be traceable, data has to stay where you put it, and a person has to remain accountable for the decision. That shapes the architecture from the first day, not as a hardening pass at the end.
We are based in Auckland and deliver for clients in New Zealand, Australia and the United Kingdom. Remote delivery is how we have always worked: the systems run in your infrastructure, in your region, wherever that is.
The commercial basics
What is in the contract
- You own the developed IP
- On payment, the IP developed for you is yours. We retain our background IP: the tooling and patterns we bring with us.
- Source-code continuity
- Every contract carries a source-code continuity clause. You keep the code, the infrastructure and the documentation, and you can hand it to anyone.
- Support is optional
- A development engagement covers the build. Afterwards we can support and maintain the system, or hand you the code, documentation and resources to run it in-house. Your call, not a lock-in.
In their words
“Silky Insights’ approach is a kind of down-to-earth and pragmatic innovation, focusing less on the theoretical potential of artificial intelligence and more on the immediate, measurable impacts it can deliver.”
“Chris and Luke are a powerhouse duo who have an amazing complementary set of skills. Already we have seen a huge difference in what we are technically delivering to our customers…”
“The team at Silky constantly evolve to fill in the gaps of our capabilities, with patience and a deep consideration for our commercial environment. We are grateful to have them, and you will be, too.”
“Silky built us a tool that has enhanced our business capability and released our staff to concentrate on value add rather than mandrolic research…”
Let’s work out what is worth building.
Thirty minutes, no deck, no discovery phase. If it is not worth building, we will say so.