One AI platform every team will actually use.
9d cowork is the agentic AI platform for organisations. Not a chatbot that answers, not an autonomous agent running behind your back. A coworker your people brief, review, and sign off on.
Self-host or managed. Your database, your identity provider, your models. Conversations are client-side encrypted before they touch the server — your data, your keys.
The organisations that win with AI won’t be the ones with the best prompts. They’ll be the ones that make delegation a habit across every function — and save their people’s agency for the calls only they can make.The 9d cowork thesis · 2026
Everyone wants adoption. Most organisations hit the same walls. Delegation-first architecture was designed to clear them.
“We scrap most AI use cases — the quality isn’t reliable enough to hand full responsibility over.”
Fully autonomous agents aren’t good enough to accept the accountability. With 9d cowork, subject-matter experts keep the agency and focus on the parts of the process where their expertise matters. The agent handles the rest — quality stays high, risk stays contained.
“Our adoption is stuck with the enthusiasts.”
Delegation is a UX everyone already understands. Brief a goal, approve an action. Skills pre-load your team’s methods so new users get expert behaviour from day one — no prompt engineering required.
“Nothing we do spans our systems.”
MCP makes cross-system work first-class. One delegation can pull from your CRM, draft in your email, update your project tracker, and post to chat — in one conversation, with approvals where they matter.
“We can’t standardise how we use AI. Who even owns it?”
Skills turn best practices into versioned, shareable units governed by OIDC groups. Ownership lives with the team that does the work — not a central AI committee that can’t keep up.
“The model landscape changes every month. How do we stay resilient?”
Pluggable LLM adapters: Anthropic, OpenAI, Google, Mistral, or self-hosted. Switch providers without migrating prompts, tools, or skills. Model-agnostic by design.
“Individual AI subscriptions are already getting expensive.”
One licence covers your team. Per-user cost tracking with quotas. Route light tasks to small models, reserve frontier models for the work that needs them. Pay for capacity, not seats.
Pick a role to watch a real flow: the agent plans, pulls from your systems, pauses for the calls only your people can make, and commits only what’s been approved.
Every piece of knowledge work decomposes into these nine capabilities. 9d cowork handles the first eight so well you stop thinking about them — and hands the ninth back to your people, every time.
Reads your docs, wikis, tickets, and inboxes. Stays grounded in what your org actually knows.
DelegableCarries context across sessions. Knows what you shipped last week without being reminded.
DelegableSurfaces patterns in data, draws connections between sources, turns noise into summaries.
DelegableRuns tools, edits files, fires API calls, closes tickets — the visible work of doing.
DelegableOrchestrates across services, teammates, and models. Keeps a plan moving in multiple places.
DelegableDrafts messages, docs, replies in your voice. Matches tone to audience and channel.
DelegableReads the room. Knows when a problem is actually a question about scope, priority, or politics.
DelegableAdjusts its approach when feedback arrives. Learns from corrections without being re-briefed.
DelegableThe call only your people can make. What ships, what doesn’t, what’s worth the tradeoff. Delegate to the agent — keep the agency.
Yours, alwaysMost AI tools measure impact in individual productivity. 9d cowork is architected to move the numbers leadership actually reports on.
Delegation is a UX everyone already understands. Skills pre-load your team’s methods, so new users get expert behaviour from day one — no prompt engineering, no training programme.
Consolidate scattered AI spend into a single licence. Per-user quotas, per-team budgets, model-level cost routing. Pay for capacity, not seats.
SSO, OIDC group-based access, tool-level permissions, full audit trails, approval workflows. AI use becomes sanctioned and traceable — no more shadow ChatGPT.
Model-agnostic by design: Anthropic, OpenAI, Google, Mistral, or self-hosted. MCP-native for any tool. Swap providers without migrating prompts, skills, or integrations.
Your infrastructure, your identity, your models, your data. 9d cowork is a platform, not a SaaS proxy.
Docker Compose stack for on-prem or VPC deployment. Or let us host it. Same platform either way — no feature splits between editions.
PostgreSQL adapter with auto-migration into your cluster, your backups, your region. Conversations are client-side encrypted before they reach it.
OIDC to Microsoft Entra ID, Google, Okta, or any compliant provider. Per-user isolation enforced end-to-end. Groups drive access to skills, servers, and quotas.
Anthropic, OpenAI (including reasoning models), Google, Mistral, or any self-hosted endpoint. Mix and match per task. No vendor lock-in at the model layer either.
Everything you need to run agentic AI as production infrastructure, not a bolt-on.
Anthropic, OpenAI, Google, Mistral, or self-hosted. Reasoning models supported. Pick the right model per task; switch without migrating prompts or skills.
AES-256-GCM in the browser before anything touches the server. Envelope encryption with unique IV per message. Your keys, not ours.
Any MCP server works. Build your own for internal systems. The agent discovers new capabilities as you add them — no platform release needed.
OIDC to Microsoft Entra ID, Google, Okta, or any compliant IdP. Per-user isolation. Groups drive access to skills, servers, and quotas.
Per-user, per-team, per-model budgets. Cache-aware pricing across input, output, reasoning, and cache reads/writes. Route light tasks to small models, reserve frontier for what needs it.
Primary model unavailable or over quota? Automatic failover to your configured chain. Exponential backoff on network errors, rate limits, and prompt-too-long.
Codify best practice once. Fork, share with OIDC groups, enforce versions across teams. A skill library that behaves like code, not copy-pasted prompts.
Every tool call, approval, and outcome recorded. Per-tool, per-conversation permissions persisted. Destructive actions always pause for explicit sanction.
Run many tasks in parallel across conversations. Browser notifications pull you back in only when an agent needs a decision.
We’ll reach out to understand your stack, security posture, and rollout plan.