GoldNet Group · UNSW AI Institute Partnership

PCM — A Multi-Purpose Human + Agentic AI Platform Built for Institutional Research

One human director. A fleet of specialised agents. SOTA reasoning, governance, and delivery. Built on the same principles as Google DeepMind's Co-Scientist — in production, not a prototype.

PM: Sabour Hosseini · GoldNet Group
Meeting: Fri 6 Jun · UNSW
Audience: Dr Sue Keay (Director, UNSW AI Institute) · Prof Ian Gibson (Deputy Dean)
See how it works ↓

1. The Opportunity

The UNSW AI Institute is building world-class AI infrastructure across 300+ academics and 50+ research groups. The challenge is not access to AI models — it is orchestration: directing AI at complex, multi-stage problems with auditability, institutional memory, and accountability.

PCM (Pantheon Co-Scientist) addresses exactly this. It is a production-grade platform where a human director commands a fleet of specialist AI agents — each with a defined role, persistent memory, and structured output — through a governed process. It is already running in production across transport safety, financial trading, and platform engineering domains.

What UNSW AI Institute needs

  • AI that scales beyond one-off prompts
  • Institutional knowledge that persists across sessions
  • Structured, auditable research processes
  • Interdisciplinary AI that bridges departments
  • Research commercialisation pathways

What PCM delivers

  • Fleet of specialist agents — each with a defined role
  • Persistent canonical memory — institutional knowledge never lost
  • 9-step structured inception — every project goes through a governed process
  • Multi-faculty deployment — Engineering, Science, Business, Law
  • Production track record — 6+ months live, multi-domain

2. What PCM Is — Not a Tool, a Team

PCM is a multi-agent orchestration platform. The human director is the accountability anchor — the person who signs off, who is answerable to boards, funding bodies, and colleagues. The fleet of specialist agents does the reasoning, research, and analysis — at scale, in parallel, with institutional memory.

Director · Human

The Director

You. Sets the problem, reviews outputs, owns the decision. Agents work for you — not the other way around.

HEPHAESTUS · PM Agent

Project Coordinator

Coordinates the fleet, tracks dependencies, manages timelines. The agent that keeps everything moving.

ATHENA · Governance Agent

Strategy & Governance

Governance, strategy, risk assessment. The agent that asks: is this the right thing to do? Is this defensible?

PROMETHEUS · Engineering Agent

Technical Architecture

Hard engineering, algorithms, data pipelines, implementation. The agent that builds the thing correctly.

APOLLO · UX Agent

Presentation & Communication

Visualisation, presentation, user experience. The agent that makes the output understandable and compelling.

THEO · Research Agent

Outside-In Research

Market intelligence, external research, competitive positioning. The agent that challenges assumptions.

Key difference from commodity AI: every output is traceable to a named agent, a canonical source, and a reasoning chain. For institutional use — where decisions need to be defended — that is not optional.
Built on GAIOS — the GoldNet AI Operating System

3. GAIOS Architecture — How It Works

The GAIOS backend is production infrastructure, not a prototype. It runs on a tiered LLM architecture with persistent memory, canonical truth stores, and a governance layer.

GAIOS Platform · VPS: ai.goldnetgroup.com.au · Pantheon Fleet HUMAN DIRECTOR You Sets direction Owns decisions HEPHAESTUS PM Agent Coordinates fleet Tracks timelines ATHENA Governance Strategy · Risk Compliance PROMETHEUS Engineering Architecture Algorithms APOLLO Presentation Visualisation UX · Render THEO Outside Voice Market research Sanity check INFERENCE PROXY · :8899 Tier cascade router Auth + auto-recall Tier-force routing Token accounting MEMORY LAYER Canon · ChromaDB · SPINE Canonical truth Vector search State graphs TIER CASCADE · :8896 Claude Relay · Multi-model T0: MiniMax-M2.7 (primary) T2: Sonnet · T4: Opus 4.7 T9: GPT-5.5 critic bridge CRITIC CHANNEL · WireGuard GPT-5.5 adversarial Orthogonal critique Gate A + Gate B All outputs vetted PRODUCTION Uptime: 6+ months 32/32 tests passing Systemd services Cloudflare CDN AUDIT TRAIL Every request logged Agent attribution Token accounting mcp_audit rows GOVERNANCE ATHENA ratification Canonical outputs Two-critic gates Human sign-off LLM UPSTREAM Multi-model platform MiniMax · Claude · Sonnet GPT-5.5 · OpenAI Auto-failover cascade DEPLOYMENT Zeus: 116.203.22.173 Caddy web server SSL · Cloudflare VPS + container — forward path — memory/enrichment — critic channel — production/governance

Multi-model fleet

Six minister agents, each potentially on a different LLM — producing genuinely diverse perspectives rather than groupthink from a single model.

Persistent memory

Canonical documents, vector search (ChromaDB), and structured state graphs (SPINE) — institutional knowledge that survives across sessions.

Adversarial critique

All significant outputs pass through two independent adversarial passes via GPT-5.5 before ratification. Bad ideas die before they reach the director.

Production-grade

6+ months live uptime. 32/32 calibration tests passing. systemd services. Cloudflare CDN. Not a prototype.

4. Co-Scientist Alignment — Why It Matters for UNSW

In May 2026, Google DeepMind published Co-Scientist (Nature) — a multi-agent system on Gemini that generates, debates, and evolves hypotheses for hard scientific problems. Co-Scientist reproduced 10 years of antibiotic resistance research in 72 hours.

The Pantheon Co-Scientist platform (PCM) implements the same principles in production — with a critical addition: a human director in the loop.

Co-Scientist (Google DeepMind)

  • Generation phase — parallel divergent hypothesis generation
  • Proximity clustering — semantic grouping of candidates
  • Reflection — virtual peer review via adversarial agent
  • Tournament — pairwise Elo ranking of hypotheses
  • Evolution — iterative refinement by recombination
  • Meta-review — system-level synthesis

PCM (Pantheon Co-Scientist)

  • Same 9-step inception protocol — fully implemented
  • Human director owns the decision — not automated
  • Different model voices per agent — more diverse than single Gemini
  • Canonical memory layer — persistent institutional knowledge
  • Production deployment — 6+ months live, not a demo
  • Gate A + B dual critique — GPT-5.5 adversarial vetting
PCM takes Co-Scientist from a research paper into a production platform — adding institutional accountability, human oversight, and persistent canonical memory. For UNSW, this means AI that can work across research projects, remember institutional context, and produce outputs that are defensible to funding bodies and review committees.

5. The 9-Step Inception Protocol

Every project on PCM goes through a structured 9-step process. This is how complex problems get decomposed, stress-tested, and resolved — without defaulting to the first idea that arrives.

Step 1
Brief intake
Director defines the problem. Task snapshot frozen. No stale assumptions.
Step 2
Generation
Each agent posts 2–3 candidate angles in their own voice. Parallel. No gate.
Step 3
Proximity clustering
Semantic grouping of candidates. Protected outlier handling.
Step 4
Reflection — Gate A
Two-critic adversarial pass. Disagreement is signal, not failure.
Step 5
Tournament
Pairwise Elo ranking per dimension. Blinded. Mechanically aggregated.
Step 6
Evolution
Top 2–3 merged and refined. Cross-pollination from multiple agents.
Step 7
Expert wrap-up
Domain expert review by role. Advisory, not authoritative.
Step 8
Director dispatch
Human director reviews, approves, and dispatches build.
Step 9
Ratify
Gate B ratification. Canonical record. Meta-review feeds next project.
Generate  Debate  Evolve  Build / Close

6. What UNSW Gets

For Dr Sue Keay — AI Institute Director

Sue Keay has 30+ years shaping Australian robotics and AI. She founded Robotics Australia Group, led Australia's robotics roadmaps, and is focused on translating AI research into practical applications. Her priority: AI infrastructure that can handle real-world complexity.

AI for interdisciplinary research

PCM can coordinate across UNSW's 50+ research groups — Engineering, Science, Business, Law, Medicine, Arts — without requiring every researcher to become an AI prompt engineer. The fleet handles the orchestration; researchers focus on domain expertise.

Research commercialisation pathway

The 9-step inception protocol structures the path from idea → validated hypothesis → research plan. PCM's adversarial critique gates mean only vetted ideas progress — reducing noise and focusing resources on high-conviction directions.

Persistent institutional memory

Research projects often span years. PCM's canonical memory layer means the platform remembers institutional context, prior decisions, and previous findings — reducing redundant work and maintaining continuity across personnel changes.

Autonomous systems AI coordination

PCM's multi-agent fleet architecture mirrors how complex robotic systems need coordinated subsystems. Applying PCM to autonomous systems research gives Sue a testbed for AI coordination at institutional scale.

For Prof Ian Gibson — Deputy Dean (Industry, Engagement, Innovation, Research)

Ian Gibson spent 30+ years in computer science and R&D management — including leading research at Canon's R&D lab and running Intersect Australia. His priority: demonstrating AI that delivers measurable industry and research outcomes.

Production-grade AI platform

This is not a research prototype. PCM has 6+ months production uptime, 32/32 calibration tests passing, systemd services, and Cloudflare CDN. The architecture is documented, auditable, and already delivering results in transport safety, financial trading, and platform engineering.

Technology shipped in 1B+ units track record

Ian's career is defined by taking research and getting it into products that ship at scale. PCM is built by an engineering team with the same orientation — the deliverable is not a report, it is a working platform that produces results.

Industry partnership model

GoldNet's engagement with UNSW is structured as a genuine partnership — Sabour Hosseini is the principal, and the PCM platform is offered for UNSW's use. This is not a vendor sale; it is a collaborative development relationship.

Research acceleration

Co-Scientist demonstrated 10 years of antibiotic resistance research reproduced in 72 hours. PCM implements the same principles for any domain. For UNSW researchers, this means machine-speed hypothesis generation and stress-testing — accelerating the pace of publishable research.

For the UNSW AI Institute — Overall

Interdisciplinary AI coordination

The Institute spans 7 faculties. PCM's multi-agent fleet can parallel-coordinate across them — each faculty research agenda gets a specialist agent perspective, unified under the human director.

Responsible AI deployment

PCM's adversarial critique gates and ATHENA governance agent mean AI recommendations are stress-tested before reaching the director. For a research institute operating under public scrutiny, this accountability structure matters.

Scalable AI infrastructure

The tier cascade architecture routes to the most cost-effective model first, scaling up only when needed. For an institute managing 300+ academics, this is not a toy — it is infrastructure that handles volume.

Commercialisation support

Research commercialisation is a stated Institute objective. PCM can support this: structured inception of commercial ideas, adversarial vetting of go-to-market strategies, persistent institutional memory of partnership discussions.

7. PCM Creation Roadmap

The platform did not appear fully formed. It went through a structured evolution — mirroring the same 9-step inception process we propose for UNSW research projects.

Completed Current Planned
2026-04 shipped

Pantheon v1 — MyClaw seats

Initial multi-seat architecture. Single-bot-per-seat model. No persistent memory. Functional but not production.

2026-05-17 shipped

Fleet redesign — self-hosted containers

Migrated to self-hosted OpenClaw containers. Named ministers with distinct roles. First identity seeds authored. Fleet PM authority scoped.

2026-05-19 shipped

Co-Scientist gap analysis — v1

Read Google DeepMind's Co-Scientist (Nature 2026-05-19). Mapped its seven specialised agents onto our six ministers. Identified five structural gaps. First proposal of 9-step inception protocol.

2026-05-29–30 shipped

9-step protocol + canonical memory

Integrated plan v1 + v2 (six load-bearing fixes). SOUL+context layer. SPINE dependency graphs. Two adversarial critique passes via GPT-5.5 PC bridge.

2026-05-31 shipped

PCS v3 — full integration live

Minister-edge gateway. Preamble auto-construct. 32/32 calibration tests pass. Full observability layer operational.

2026-06-01 shipped

ENFORCE mode + PCS Heartbeat v2 LIVE

PCS Heartbeat v2 spec critic-vetted. Automated state monitoring across all six minister agents. DRY_RUN soak active. System fully hardened for production multi-domain operation.

2026-06-02 current

UNSW Partnership — First deployment

PCM offered to UNSW AI Institute as a collaborative development partnership. Live platform demo, institutional onboarding design, and first research project scoping.

Ready to discuss — Fri 6 Jun, 10:00, UNSW

Dr Sue Keay (Director, UNSW AI Institute) · Prof Ian Gibson (Deputy Dean)
Sabour Hosseini · GoldNet Group

Platform live: analysis.goldnetgroup.com.au · trading.goldnetgroup.com.au · vicrisk.goldnetgroup.com.au