Response to the Wiseway AI Brief

Build on standards, not platforms.

An architecture for Wiseway's first two AI assistants — the Knowledge Assistant and the Sales Quoting Assistant — that delivers fast value today and is unlikely to need rebuilding as the AI industry evolves.

First pilotSix weeks · HR Knowledge Assistant
Second pilotTwelve weeks · Sales Quoting Assistant
Data residencyAustralia, end-to-end
The approach

A foundation made of five small parts.

Rather than picking a single vendor — Microsoft, Databricks, Snowflake — and accepting whatever they offer, we build on the open standards the AI industry agreed upon in late 2025. Five small components, each replaceable, together delivering one coherent platform. Click any card to see what it does.

01
The brain
Claude, the AI model
02
The operating system
Claude Agent SDK
03
The playbook
Skills in Wiseway's git
04
The plugs
MCP connectors
05
The flight recorder
OpenTelemetry GenAI
The architecture

How the pieces fit together.

Click any box to see what it does, in plain English. The shape repeats for every future agent — Customer Service, AP, AR, Reporting — so you scale by writing new playbooks on the same foundation, not by adding new platforms.

Every action above is recorded by the flight recorder (OpenTelemetry) and audited in Microsoft Purview.
Core runtime
Open standard
System / connector
The two agents

What we'd build first.

Two agents, one foundation. The Knowledge Assistant proves the architecture is real; the Sales Quoting Assistant proves it is commercially valuable. Both reuse the same connectors, the same identity flow, the same audit trail.

Agent 01

Knowledge Assistant

Staff ask HR or SOP questions in Teams, by email, or on the web. The assistant retrieves the right policies from SharePoint, answers in plain English, and cites the source document.

The safety property that matters: the assistant only ever sees documents the person asking is already entitled to see. If you remove someone from a permissions group on Monday, they immediately stop getting answers from documents they no longer have rights to.

Pilot: one HR library, AU staff only, six weeks. Expand to all jurisdictions and document drafting in a further six weeks.
Agent 02

Sales Quoting Assistant

An RFQ arrives. The agent reads it, looks up current rates and surcharges from a governed rate-card store, applies client-specific terms, and drafts a quote in your branded format.

The rep always sends. Below-margin quotes are flagged for manager approval in Teams. Every rate used and every approval is logged with full audit trail — the SHEIN-style wrong-rate scenario is prevented by design, not by good intentions.

Pilot: one division (e.g. Air Freight), one rep team, four weeks once the rate-card store is in place.
Why this lasts

The same bet that won the internet.

In the last eighteen months the AI industry has agreed on a small set of open standards — the way it once agreed on Wi-Fi, SQL, and HTTP. The most important one was donated to the Linux Foundation in December 2025, and is now backed by every major AI lab and cloud provider. History tells us that when this kind of consolidation happens, the open standard wins the long run, and proprietary platforms eventually rebuild around it.

Building on standards means…

  • You can change AI models — Claude now, a better one in three years — by changing one line of configuration. Your agents don't need to be rewritten.
  • You can change data platforms — PostgreSQL now, Snowflake or Databricks later — by swapping the connector behind the scenes. Your agents don't change.
  • You can change cloud providers if Australian sovereignty rules tighten. The standards work on Microsoft, Amazon, and Google equally.
  • You own the intellectual property of every agent — its instructions, its connectors, its evaluations — as Markdown and code in a Wiseway repository, not locked inside any vendor's console.
What we ask Wiseway to commit to

Three things, in order.

  1. The intellectual property lives in Wiseway's code repository.
    Agents, skills, prompts, evaluations, connectors — all of it. The consultant builds the first version; Wiseway owns and extends it from day one.
  2. A six-week first pilot of the Knowledge Assistant.
    One HR library, AU staff only. Prove value with real users before expanding. Brief said it best: pilot small, prove value, expand.
  3. Open standards as the default.
    Any tool or library we propose must work on the open standards (MCP, A2A, OpenTelemetry). Any exception will be explicitly called out so the trade-off is conscious — never accidental.