Fixed-Price AI Workflow Services – No Hourly Billing, No Enterprise Overhead
Auburn AI offers three fixed-scope service tiers, all priced in Canadian dollars. You know the cost before work starts. There are no change-order surprises, no retainer commitments, and no account managers between you and the person doing the work.
Tier 1 – AI Workflow Audit ($497 CAD)
Not sure where automation actually fits in your business? This is where to start. Over one week, Alexander reviews how your team currently moves information – what tools you use, where handoffs happen, where things fall through the cracks.
You get a written deliverable with three specific workflow candidates, ranked by effort-to-impact. Each candidate includes a plain description of what it would do, what tools it would touch, and a rough complexity rating. No commitment to build anything is implied or expected.
- Timeline: 1 week from intake call
- Deliverable: Written audit report, three automation candidates
- Good fit: You know AI should be doing something in your business but you have not pinpointed what
Tier 2 – Single Workflow Build ($2,500 – $5,000 CAD)
One production-ready automation, built and deployed in your environment. The price range reflects complexity – a straightforward email-to-task workflow sits at the low end; a multi-step agent that reads documents, reasons about them, and routes output sits higher.
Every build includes testing against real data before handoff, written documentation your team can actually follow, and ten days of post-launch support for issues that surface in normal use.
- Timeline: 2 to 3 weeks
- Deliverable: One live workflow, documented, tested
- Post-launch: 10-day support window via email
- Good fit: You have a specific, painful manual process and want it solved cleanly
Tier 3 – Full Stack Setup ($7,500 – $15,000 CAD)
For businesses that want a proper AI workflow foundation, not just a single fix. This tier covers installing and configuring a self-hosted n8n instance, building three to five interconnected workflows, and standing up the infrastructure so your team can maintain and extend things after the engagement ends.
The 30-day support window gives you room to run the system through a real month of operations before you are on your own.
- Timeline: 4 to 5 weeks
- Deliverable: n8n instance, 3 to 5 production workflows, documentation, team walkthrough
- Post-launch: 30-day support window via email
- Good fit: You are ready to commit to AI-assisted operations and want it done right the first time
What These Workflows Actually Look Like
Here are five concrete examples of what gets built under these tiers. These are not theoretical – they reflect the kinds of problems Canadian small businesses actually bring.
- Inbox triage: Incoming emails are read by Claude, classified by urgency and type, and routed to the right folder or person with a one-line summary attached. Your inbox stops being a to-do list you sort manually.
- Proposal drafter: A intake form or CRM trigger kicks off a workflow that pulls client context, runs it through a Claude prompt built around your service offering, and outputs a first-draft proposal into Google Docs or Notion for your review.
- KPI digest: Every Monday morning, a workflow pulls numbers from your tools – sales data, support tickets, ad spend – formats them into a short summary, and sends it to your inbox or Slack channel. No dashboards to check.
- Meeting-to-actions: A transcript from Zoom, Teams, or Otter drops into a folder. The workflow reads it, extracts action items with owners and deadlines, and creates tasks in your project management tool automatically.
- Client status updates: A trigger fires when a project milestone changes in your tool of choice. The workflow drafts a status update email in your voice, ready for you to review and send in under a minute.
What Is Not Included
Being clear about scope saves everyone time.
- Enterprise procurement: Auburn AI is a sole proprietorship. Vendor registration, security questionnaires for Fortune 500 suppliers, and multi-signature contracts are not a fit.
- AI strategy decks: If you need a presentation for your board explaining AI adoption, that is a consulting product different from what is offered here.
- Custom AI products: Building a customer-facing SaaS tool or a trained model is product development work, not workflow automation. That is out of scope.
- Real-time voice AI: Live phone agents and real-time speech processing require infrastructure and latency management that falls outside these service tiers.
Tools and Stack
Every engagement uses a short, maintainable stack. No proprietary platforms that lock you in.
- n8n: Self-hosted on a Hostinger VPS or on your existing infrastructure, or n8n Cloud if self-hosting is not appropriate. All workflow logic lives in a place you control.
- Claude API (Anthropic): Used for any step that requires reading, reasoning, drafting, or classifying. Model selection depends on task complexity and your cost tolerance.
- Your existing tools: Google Workspace, Microsoft 365, Notion, HubSpot, Slack, and most tools with a REST API or webhook connect without issue.
- PIPEDA-aware architecture: Data handling is designed with Canadian privacy law in mind. Where possible, data stays in Canadian or client-controlled infrastructure. Prompts sent to the Claude API are scoped to contain only what is necessary for the task.
How to Start
Send one sentence describing your problem to alexander@auburnai.ca. Something like “We spend two hours every week manually copying data from emails into our spreadsheet” is enough to get a useful reply. If it sounds like a fit, Alexander will schedule a short call to confirm scope and quote the right tier.
No intake form, no sales funnel, no automated response sequence. Just a direct email to the person who will do the work.
Have a process you would rather not run by hand anymore? Email Alexander.