Services

Start small. Prove it. Then build.

Every engagement starts with a question about your business. The steps below are how we earn the right to build.

One day · fixed fee

Design your data workshop

A day that gives your data a design. We score the questions your business runs on, six tests each, and cost what it takes to make them ready. $18,000 plus GST, fixed, and credited in full to a Data proof signed within 60 days of the workshop.

What happens on the day →
Two to three weeks · fixed fee, quoted after the workshop

Data proof

Proof on your own data: your top question answered by a Cortex Agent in your account, whether your customer signals join, and a clear go or no-go before any platform decision.

Build

Your tenancy. Your data. We build the engine.

Snowflake as the data plane, an agent layer on top. Your data is stored in Australia, in your own Snowflake account. Where a model runs outside Australia, we tell you first and your admin decides. Patterns we have built: AI concierge and voice for multi-site operators, intelligence nodes with a 6am brief, retrieval for frontline workforces, national data-collection platforms, research and knowledge-sharing platforms for government and health.

How we build on Snowflake →
Run

Silk+

Keep it secure, keep it moving. Support plans with service levels, security patching and ongoing evolution, all with Australian data sovereignty.

What we build on Snowflake

Six things clients ask us for.

We build each one in your own Snowflake account.

Agents your people can ask

Questions answered from your own data in plain words, with the working shown.

Built with
Cortex Agents
Semantic views
Snowflake CoWork

The 6am brief

A short note to each owner before the day starts: what changed overnight, and what needs attention.

Built with
Tasks
Semantic views
Cortex AI Functions

One customer across every channel

Store, online, loyalty and franchisee records matched by rules we agree with you, so an order, an offer and a complaint land on the same person. Each franchisee sees their own stores' customers, and no one else's.

Built with
Snowflake Openflow
Dynamic tables
Masking and row access policies

Voice and concierge

An agent layer under your voice and chat. A verified customer can check their own order, booking or membership. Your MCP server confirms who they are, and the channel's role sees only what that channel needs.

Built with
MCP server
Cortex Agents
Masking and row access policies

Documents and frontline knowledge

Fields pulled from forms and letters, and answers for frontline staff with the source cited.

Built with
Cortex AI Functions
Cortex Search
Streamlit in Snowflake

Collection and research platforms

National data-collection and knowledge-sharing platforms, so researchers, agencies and clinicians work from the same data.

Built with
Secure Data Sharing
Horizon Catalog
Snowpark
The workshop

One day. Your people, our method, a design for your data.

You bring the people who hold a piece of the customer. We bring the designers and a data architect. Fixed fee, and you leave with something you can act on the next morning.

The question the workshop starts with

Which of your numbers would you trust an agent with?

Morning

Your questions, and where each answer lives.

Stores, online, loyalty, franchisees, service, finance. We map every place a customer leaves a trace.

Afternoon

The joins, a live test, the first three wins, costed.

Where one person shows up as several records. Then one of your people asks an agreed question in their own words, and we check the answer against a number you trust.

You leave with

One pack, in three business days.

A question register. A signal map. A readiness score for each question on six tests: defined, joined, fresh, governed, trusted, tested. A target design. A 90-day plan, and a memo for your executive.

Who should be in the room

Six to ten people. Whoever owns the customer, whoever owns marketing, whoever owns the data, and one voice from a store, a franchise or the front line.

Who runs it

Tim Evans, Design Director, and Melissa Lau, Design Lead, with a dptr data architect and a dptr Snowflake engineer. The people who run the day are the people who would build what comes next.

Not sure where to start? Start with the workshop.