What is DM?
DM is shared context for human + AI teams, built by AuzzurA. It keeps Rules, Skills, Decisions, and repository intelligence with the work so the next person or connected agent can pick up with less reconstruction. Memory is on the same product path and is in progress. It is not a meeting note, not a chatbot, and not a Mem0 or Graphify wrapper.
Does DM include memory and repository intelligence?
Repository intelligence is ready: code structure, dependencies, relationships, and architecture context around the task, with Graphify as the current integration. Memory (Mem0 integration) is in progress on the same product path. Neither provider is the product.
Is DM a meeting note-taker?
No. Notes and transcripts are evidence, not the product. DM preserves what was officially decided, why, under which assumptions, what evidence supported it, and whether it is still current. Meetings are one evidence source among many.
How is this different from enterprise search or RAG?
Search and RAG retrieve from what already exists. DM keeps what the team decided current, shared, and approved — and brings the subset that matters for this task to the next person or connected agent. Memory is a start; recall alone is not the whole product.
Does AI automatically create official knowledge?
No. AI proposes candidates. A human reviewer edits, rejects, requests more evidence, or approves. Nothing becomes trusted organizational knowledge without human approval — a core design principle.
Does DM replace Jira, Confluence, or documentation tools?
No. Those tools remain where teams plan, document, and execute work. DM is the shared context across them: what was decided, why, whether it is still current, and what the next person or agent needs.
Does DM crawl our entire workspace?
No. DM operates on evidence users intentionally add or sources explicitly configured for the workflow — a transcript, document, uploaded file, ticket, email thread, chat excerpt, or project artifact. There is no background crawl of the whole company workspace.
What does an AI agent get access to?
Only the context they are allowed to use that applies to the task — not the entire workspace, and not every rule file in every repo.
Can we keep data in our own environment?
Managed DM is available now, with Control and Knowledge data in physically separate, workspace-bound stores. A dedicated or customer-controlled Knowledge Store (BYODB, Enterprise) is provisioned with AuzzurA during onboarding. Bring Your Own Key and Bring Your Own Model are planned, not available yet. See Trust & security for the full maturity model.
Is DM GDPR compliant?
We do not claim certified or audited GDPR compliance. DM is designed around data-protection principles such as minimization, purpose limitation, scoped access, and customer-defined data boundaries. Hosting, residency, retention, subprocessors, and other legal/technical requirements are agreed as part of your deployment; formal compliance claims match current legal and technical documentation.
Does DM train models on our data?
We do not publish a blanket model-training guarantee here. Provider configuration and retention are agreed as part of your deployment. Bring Your Own Model (your own provider or approved endpoint) is planned, not available yet.
Who is it for?
The first focus is product, engineering, architecture, platform, and AI-enabled teams — including the coding agents those teams already run. Memory users, repository-context users, and human + AI teams are all in the door. Coding agents are the concrete beachhead, not the whole product.
What value does DM deliver?
Value can be measured by decision recall, context reconstruction effort, rework from missing constraints, re-litigated decisions, conflicts caught before implementation, onboarding/handover time, and reuse of approved knowledge. Exact success criteria are agreed case by case — we do not publish invented benchmark numbers.
Is DM publicly available?
Early access is available by request. We do not offer a public self-serve Start Free signup in this release. Scope, terms, and data handling for a team are agreed when we follow up.
What does getting started with DM look like?
Onboarding is scoped to one team and one workflow or decision domain to start. We agree upfront what evidence can be used, what success means, and which workflow should be tested first — for example coding-agent alignment, architecture decisions, or leadership handoff.
Your team already made the decision. Make sure it stays made.
DM preserves the approved rationale, evidence, owner, constraints, and change history behind important work — without replacing the tools your team already uses. Onboarding is scoped and discussed case by case with your team.