01 · Approved knowledge
Only what you’ve approved
Each Aveya application draws from a defined source base you control — not the open web, and not everything you own.
Governed organisational AI
A governed AI product, grounded in your organisation’s knowledge.
Create source-backed answers from approved knowledge, with controlled access, visible evidence, and oversight built in.
Configured around your approved knowledge, with private and isolated deployment options available where required.
Only approved sources
Approved knowledge
Policies & procedures
Contracts
Operational records
Approved guidance
Governed AI
Aveya
Scoped · access-controlled
◦ private deployment options
Source-backed answer
“Where did Q2 renewal risk land?”strict mode
Three mid-market accounts, ~$840K at risk.
Delayed onboarding1 and two open security reviews2.
The problem, plainly
General AI will answer anything. Your organisation needs answers it can stand behind.
The hard part isn’t getting an answer. It’s knowing it came from approved sources, that the right people can see it, and that you can show the evidence.
The governed layer
Approved knowledge in. Source-backed answers out.
Aveya is the organisation-specific layer of governance around your AI: approved sources, controlled access, evidence and oversight.
02 · Evidence
Answers show their sources
Answers show the approved sources used to support them. In strict mode, unsupported answers are withheld rather than presented as grounded.
03 · Control
Access & oversight
Scoped by tenant, application, assistant and role; retrieval fails closed. Key activity is retained for review and operational oversight.
A different deployment model
Aveya is the second path.
One path is to commission an AI system around a specific engagement. Aveya is the other: deploy an existing governed product and configure it around your organisation’s approved knowledge, access model and operating requirements.
Custom system
- Designed and assembled for a specific engagement
- Infrastructure, ingestion, retrieval and governance form part of the build
- Ongoing platform ownership depends on the delivery arrangement
Aveya
- An existing governed product, deployed through the agreed operating model
- Configured around approved knowledge, access and organisational requirements
- Retrieval, evidence, oversight and continued product development remain part of Aveya
Inside Aveya
Ask. See the evidence. Retain what needs review.
Questions in plain language, answered from the approved source base, retrieval scoped to your tenant and application. When an output needs review or handover, keep it with its question, sources and revisions.
Where did Q2 renewal risk land?grounded
Risk sits in three mid-market accounts.
Combined ARR at risk is ~$840K, from delayed onboarding1 and two open security reviews2. Northwind is highest priority after last week’s escalation3.
Sources · 3
1
Onboarding tracker
Approved · 2d ago
2
Security review log
Approved · Fri
3
Northwind escalation
Approved · Jul 3
Operational intelligence
Built for the work your organisation runs on.
Aveya is most valuable where answers need to be reliable, source-backed and governed.
Lead Conversion Intelligence
Lead Conversion Intelligence connects lead source, campaign activity, follow-up and sales outcomes into one evidence-backed view, so teams can review what happened after the lead arrived.
Medical Practice Operations
Aveya gives practice teams a governed application for billing, compliance, correspondence and policy questions. Answers are source-backed, reviewable and kept with the context needed for later use.
School Operations Intelligence
Aveya gives staff source-backed answers for leave, enrolment, onboarding, compliance and administrative questions from approved school policies. Operational knowledge only. Never teaching, assessment or student-facing use.
Deployment & trust
Private deployment options, where control matters.
Aveya can run as an Aveya-managed environment or, for qualifying engagements, as a private/customer-controlled Azure deployment.
Every retrieval is restricted to your tenant and approved workspace, and fails closed.
Runs on Azure OpenAI; under Microsoft’s terms, your prompts and completions aren’t used to train the models.
Authentication fails closed: Aveya won’t start in production with unsafe auth configuration.
Region and residency requirements are agreed as part of the deployment model.
◦ private deployment options
Approved knowledge
Defined source base
Controlled applications
Access by role
Aveya
Governed retrieval & source-backed answers
retrieval stays scoped to your tenant · fails closed
Two ways in
Run it directly, or build on it.
Run Aveya directly
Governed AI on your approved knowledge, with the access, evidence and oversight your organisation requires.
- Answers from approved sources, with visible evidence
- Access controlled by tenant, application and role
- Private or customer-controlled deployment available where required
Build Aveya into your solution
The governed intelligence layer underneath selected partner-led solutions, where governance, privacy, evidence or deployment control matter.
- Sits under your delivery; needn’t power every engagement
- Customer-controlled Azure deployment available for qualifying engagements
- Source-backed answers your client can review
See governed organisational AI in practice.
See Aveya answer from approved organisational knowledge, with visible sources, controlled access and evidence attached.
20–30 min · practical walkthrough