All customer stories

Deploying secure, multi-model AI at scale without technical debt

How Nous Group, a leading management consulting firm operating across Australia, Canada and the UK, gave every consultant access to governed AI inside their Azure environment, without building and maintaining their own platform.

NG
Nous Group
Australia, Canada & UKManagement Consulting

“We wanted a platform that gave staff advanced AI tools without creating technical debt.”

David Diviny, Nous Group

Overview

Nous Group is a leading management consulting firm with offices across Australia, Canada and the UK, competing with the Big Four, McKinsey and BCG across health, education, human services, energy and decarbonisation, defence and other complex public and private sectors.

As demand for AI-enabled consulting grew, Nous wanted to give every consultant access to advanced AI tools without building and maintaining a bespoke internal AI platform, taking on long-term technical debt, or locking into a single AI vendor.

SmartSpace was selected as the AI platform to sit inside Nous' own Azure environment, orchestrating multiple frontier models and complex workflows, and surfacing them directly in Microsoft Teams.

The challenge

Executives

  • Enable AI for all staff fast, without hiring a specialist AI team
  • Keep total cost of ownership predictable and sustainable
  • Stay ahead of competitors while reassuring boards on risk, privacy and IP

Governance and Risk Leaders

  • Prevent consultants using uncontrolled public AI tools with client data
  • Ensure AI runs in a governed environment aligned with existing Azure policies
  • Maintain clear auditability of where data goes and which models are used

Technology leaders

  • Provide a single platform for multi-step workflows, not just chatbots
  • Avoid vendor lock-in to a single model such as Copilot or ChatGPT
  • Integrate cleanly with Azure security controls and collaboration tools

Delivery teams

  • Move beyond a one-off Azure chatbot that is hard to evolve
  • Build and iterate on complex workflows without standing up new infrastructure
  • Enable non-developers to use AI safely while retaining full code-level control

The SmartSpace solution

Nous deployed SmartSpace as a self-hosted AI orchestration platform inside their own Azure subscription. SmartSpace provides:

1

Workspaces per use case

Encapsulating data, models, access controls and tools behind a clean API for each business need.

2

Data connectors and Data Spaces

Continuously ingest and vectorise content from Azure and other systems, ready for retrieval-augmented generation.

3

Model-agnostic layer

Switch between Gemini, Claude, ChatGPT and others per workspace, without re-architecting solutions.

4

Repurposable UI in Microsoft Teams

AI tools appear in the flow of work for consultants, with no new system to learn or adopt.

5

Enterprise-grade security and governance

Backed by a continuously monitored ISMS, strong encryption, and Drata-verified controls for access, logging, backups and incident response.

How Nous uses SmartSpace today

01

Company-wide internal AI chatbot

A single chatbot accessed via Teams, available to all staff. Built on SmartSpace Workspaces and Data Spaces, so it can safely use Nous' internal content and be iterated over time without rebuilding infrastructure.

02

Client-specific proof of concept with complex workflows

A bespoke, multi-step workflow for a specific client, using SmartSpace's workflow framework to combine document search, reasoning steps and custom tools. Designed as a POC that moves straight to production because it already runs in Nous' Azure environment.

03

Foundation for future internal tools

With SmartSpace deployed to all staff and embedded in Teams, Nous can add new chats and workflows quickly, reusing the same platform, data and security guardrails without starting from scratch.

Outcomes

Executives

  • Strategic clarity: SmartSpace clarified that staff should use internal, governed AI tools rather than public ones for client work, reducing unmanaged risk and signalling maturity to boards and clients
  • Faster time from idea to impact: new ideas can be prototyped on SmartSpace in days and moved straight into production, avoiding the typical POC that never ships
  • Cost-effective adoption: platform-level pricing with no need to stand up new infrastructure for each use case means Nous can scale usage without cost scaling linearly with headcount

Governance and Risk Leaders

  • AI governance that speeds things up, not slows them down: running SmartSpace in Nous' own Azure tenant gives confidence that AI use is auditable and aligned with emerging governance expectations, including ISO 42001-style controls and IoD guidance on AI oversight
  • Trusted environment for frontier models: consultants can choose from multiple models, but always within a secure, logged environment where data protection, privacy and IP are respected

Technology Decision Makers

  • No new technical debt: SmartSpace provides the orchestration engine, connectors, vector stores and workflow framework so the internal team focuses on use cases, not infrastructure
  • Azure-native deployment: SmartSpace is deployed from the Azure Marketplace as a managed application into the customer's subscription, inheriting Azure's security and compliance stack while allowing full customisation
  • Model and vendor flexibility: the organisation can adopt new LLMs or domain-specific models over time with minimal change to downstream tools and workflows

Architects and Delivery Teams

  • A single platform for many use cases: workspaces and workflows let architects define reusable building blocks such as document search, multi-agent reasoning and custom tools that can be wired into different projects without re-plumbing
  • Developer-friendly, low-code where it helps: SmartSpace combines full API and SDK access with admin tools so non-technical users can configure simple use cases, while engineers handle complex flows via code
  • Embedded in Teams: surfacing SmartSpace workspaces in Teams means consultants do not need to learn a new system as AI shows up where they already collaborate
  • Given every consultant access to secure, multi-model AI directly inside Teams
  • Avoided building and maintaining its own AI application stack
  • Reduced AI risk by keeping data and models within its Azure environment
  • Created a clear internal message: use the governed internal platform, not public tools
  • Built a foundation where client-specific POCs can be taken straight to production rather than dying as experiments

This is AI that works the way consulting firms actually operate: secure by design, flexible by default, and ready to grow with the business.

See how other organisations are using SmartSpace

View all customer stories