Private Business AI Assistant
A private, self-hosted AI assistant for your team - shared across staff, wired into the tools your business already uses, with the access controls and data governance to keep it safe.
Why not Microsoft Copilot or ChatGPT Team?
Per-seat AI subscriptions are generic, send your data to someone else's cloud, and bill every employee every month. This assistant runs on your infrastructure, learns your business from your own documents, and is scoped to each person's role. Cloud models stay available as an opt-in fallback for heavier tasks, on your terms.
You own your data
Runs on your hardware. No training on your data, no provider lock-in.
One cost, your whole team
A setup fee, not a monthly bill for every employee.
Scoped per person
Role-based access and audit logs. Cloud is opt-in per task.
What's Included
One-time fee. Pick the tier that fits your team.
Business AI Assistant (Basic)
A private business assistant, set up and integrated.
- Everything in the Personal assistant
- Team / multi-user access with per-user permissions
- Business integrations: Slack, Microsoft 365 / Google Workspace, CRM, ticketing, webhooks
- Shared + per-user memory and identity files
- Up to 3 business workflows configured live
- 90-minute team walkthrough + 30 days of tuning
Business AI Assistant (Advanced)
Everything in Basic, plus hardware, custom builds, and training.
- Everything in Basic
- Advanced consultation including hardware for your local setup
- Custom builds and custom skills for your workflows
- 1 employee training session included
- Compliance & governance: audit logs, data residency, retention rules
- On-prem / server deployment + priority SLA
Managed
A multi-month engagement we set up and look after.
- Everything in Advanced
- Multi-month engagement: phased setup, improvement, and testing
- Ongoing knowledge-base build-out and management
- Team training and change-management support
- Dedicated support with a priority response window
- Custom scope, priced to your requirements
Four Pillars
Team access, security, integrations, and workflows - what makes this assistant your company's.
Team Access & Control
One assistant, scoped per person.
- Shared across your team with per-user permissions
- Role-based access to data and actions
- Central admin for onboarding and offboarding
- Per-user memory plus shared company knowledge
Security & Governance
Runs on your infrastructure, on your terms.
- Self-hosted or on-prem; local by default
- Audit logs of what it accessed and did
- Data residency and retention rules
- HIPAA / legal handling where required
Business Integrations
Plugs into the systems you already run.
- Slack, Microsoft 365, and Google Workspace
- CRM, help desk, and accounting
- Knowledge base (Notion, Confluence, SharePoint)
- Webhooks and internal APIs
Workflows & Automation
Repeatable automations your team can trust.
- Team workflows configured during setup
- Permission-gated actions with human approval
- Pre-approve recurring actions for autonomy
- Ongoing improvement during the engagement
Example Workflows
Examples of what we wire up for teams. The first few are configured during setup; more are added during the engagement.
New-lead triage
Watches a shared inbox or CRM, drafts a first reply, scores priority, and pings the team in Slack or Teams.
Meeting notes to CRM
Transcribes calls, summarizes decisions, and pushes action items and updates into your CRM or Notion.
Support-reply drafting
Drafts ticket responses from your knowledge base for staff to review and approve before sending.
Daily ops brief
Posts pipeline, calendar, overdue tasks, and KPIs to a Slack or Teams channel every morning.
Document & quote generation
Fills proposals, quotes, and SOWs from your templates and CRM data, ready for a final check.
Onboarding assistant
Answers new-hire questions from your SOPs and wiki, and runs onboarding checklists.
How the Privacy Lanes Work
Two lanes, one master switch. Flip on the fly per session, per task, or per workflow.
Local Lane
Model runs on your machine. No data leaves your hardware.
- • Ollama / LM Studio / your platform
- • Sees only granted files
- • No internet required to function
Cloud Lane
Routes to OpenAI / Anthropic / your choice when enabled per task.
- • Use for heavier reasoning
- • Off by default
- • Toggle per prompt or workflow
Internet Toggle
Master switch over both lanes. When off, assistant is airgapped.
- • One toggle, no restart
- • Only allowed integrations stay live
- • Use for sensitive sessions
Hardware Requirements
Business sizing scales with how many people use it at once and what it does. Cloud fallback can absorb spikes, so local hardware is sized to typical load, not worst case.
| Scale | Concurrent users | Suggested | Notes |
|---|---|---|---|
| Small team | 1-5, mostly sequential | 1x 16-24GB GPU, 32-64GB RAM, 1TB NVMe | Comfortable for an 8-14B model + RAG |
| Growing | 6-15, frequent use | 24-48GB VRAM (or 2x 24GB), 64-128GB RAM, 2TB NVMe | Batching (vLLM); dedicated always-on server |
| Larger / heavy | 16-50, concurrent + agents | 2-4x 24-48GB or 1x 48-80GB (A6000 / L40S-class), 128-256GB RAM, NVMe RAID | On-prem or colocated server |
Your business assistant plugs into the systems your team already runs on. Every integration is opt-in, permission-gated, and audit-logged - it only does what you allow, scoped per user.
| Integration | How we connect it | What it enables |
|---|---|---|
| Microsoft 365 | Graph API (OAuth) | Teams chat, Outlook mail, SharePoint docs, and calendars |
| Google Workspace | OAuth | Gmail, Calendar, Drive, and shared docs |
| Slack | Slack app + bot token | Team channels, approvals, and daily briefs |
| CRM (HubSpot, Salesforce, Pipedrive) | API key / OAuth | Read and update contacts, deals, and pipeline |
| Help desk (Zendesk, Freshdesk, Intercom) | API token | Triage, tag, and draft ticket replies |
| Accounting (QuickBooks, Xero) | OAuth | Invoice status and AP/AR reminders (read + draft) |
| Knowledge base (Notion, Confluence, SharePoint) | API + indexing (RAG) | Answer from your SOPs, docs, and wikis |
| Webhooks / internal APIs / databases | HTTP / SQL, read-only by default | Trigger actions and pull from internal systems |
The Setup Process
Discovery to rollout, scoped to your team. Timeline depends on integrations and how much knowledge we wire in.
Discovery & scoping
We map your team, tools, data sources, and the tasks worth automating, then confirm scope and fit before booking.
Pilot setup
The assistant is stood up on your infrastructure, connected to a first integration and a slice of your knowledge base, with per-user access.
Rollout
Workflows, integrations, and the full knowledge base are wired in. Team training and permissions are configured.
Improve & support
Tuning, new workflows, and monitoring. Basic and Advanced include a tuning window; Managed is an ongoing engagement.
Runners like Ollama or LM Studio load and run the model. Assistant platforms like Hermes, Goose, and OpenClaw sit on top and add the memory, scheduling, and integrations below. Most setups use both - we pick the combination that fits your goals and hardware.
Tap any platform name for a quick description of where it came from and what it is good at.
| Capability | |||||
|---|---|---|---|---|---|
| Type | Omnichannel agent | Coding & automation agent | Local-first agent | Knowledge assistant | Model runner |
| License | MIT | Apache-2.0 | MIT | AGPL | Free, proprietary |
| Runs 100% on your hardware | |||||
| Chat apps (Telegram / Discord) | |||||
| Slack | |||||
| Calendar | |||||
| Files & notes (Obsidian vaults) | |||||
| Scheduled / unattended jobs | |||||
| Persistent memory & skills | |||||
| Built-in local model serving | |||||
| Best for | Living in your messaging apps | Dev workflows, automation, and MCP tooling | Hands-on automation that takes actions | Searching and acting on your notes | Simple, fast local model running |
These are examples, not an exhaustive list, and this space moves fast - the best harness and methods change with new releases. We re-assess and recommend the right fit during setup and your initial consultation. We are not affiliated with these projects.
The platform is the visible layer; these methods do the heavy lifting underneath. We pick and tune them to your needs.
- RAG over your knowledge base, backed by a vector database (Qdrant / Weaviate), so answers cite your own docs
- MCP and APIs to connect your business tools, with role-based access and audit logging
- Local by default with opt-in cloud routing, plus data residency and retention controls
- Team-grade platforms (Open WebUI, AnythingLLM, Dify) where a shared interface fits
- Evaluations and monitoring so quality holds as usage grows
Need ongoing help after setup?
Ask about ongoing AI support and management for your team - updates, improvements, and new workflows on a recurring basis. Talk to us →
You Might Also Like
Email, infrastructure, and the IT foundation your business runs on.
Learn MoreTrain your team to get the most out of the assistant and your AI tools.
Learn MoreNeed a server to run it on? We build hardware tuned for AI workloads.
Learn MoreFrequently Asked Questions
Ownership and cost. It runs on infrastructure you control, learns your business from your own documents, and is scoped per role. You pay a setup fee instead of a per-seat monthly bill, and your data is not used to train anyone else's model.
Almost certainly not. Enterprise stacks like Databricks Agent Bricks are built for large companies with data teams, lakehouse infrastructure, and consumption-based contracts. For a small or mid-sized team that is more platform, cost, and complexity than you need. We deliver the same kind of private, governed assistant right-sized for your business: one setup fee, running on hardware you own, with no data team required.
Yes. It is a shared assistant with per-user permissions and memory, plus shared company knowledge. Access is added or removed as staff changes, with role-based controls.
It runs on hardware you control, local by default. We set granular access, audit logging, and data residency and retention rules, with HIPAA or legal handling where required. Cloud models are opt-in per task.
Slack, Microsoft 365, Google Workspace, CRM, help desk, accounting, and your knowledge base (Notion, Confluence, SharePoint), plus webhooks and internal APIs. Every integration is opt-in and permission-gated.
It depends on how many people use it at once and the tasks it runs - see the table above. No server yet? We can spec and build one, or start on your existing hardware with cloud fallback and scale up as usage grows.
Basic ($500) and Advanced ($1,000) are one-time setups. Managed is a custom multi-month engagement (contact us). Ongoing upkeep and improvement can be added separately.
Yes, but permission-gated: it asks before sending, editing, or changing anything, and you can pre-approve specific recurring actions. You decide what runs autonomously and what needs a human.
Ready to build your assistant?
Tell us about your workflows, hardware, and the integrations you care about. We'll confirm fit and timeline before kickoff.