From Free Audit to a Production AI System in Four Stages
Every Kernel Flow build follows the same sequence. The audit and blueprint take 1 to 2 weeks and cost nothing. The build and deployment typically take 4 to 8 weeks, and you approve the plan and the price before any payment is due.
The sprint, stage by stage
- [ Stage 01 ]Week 0
Audit
Map how each team works today: the tools, the handoffs and the steps someone does by hand every week. The audit is free and carries no obligation.
- Walkthrough of workflows, systems and data sources
- List of the processes where AI returns the most
- Access requirements for the build, if you go ahead
- [ Stage 02 ]Weeks 1 to 2
Blueprint
Receive a written, team-by-team blueprint showing what to build, which systems it connects to and the exact price. Measurable KPIs are documented and signed off by both sides before any build starts.
- System architecture and integration map
- Fixed scope, fixed price and signed KPIs
- 40% payment only once you approve it. Reject it and nothing is owed
- [ Stage 03 ]Build: 4 to 8 weeks
Sandbox Build & ERP Wiring
Engineer the system against sample, anonymised or synthetic data first, then wire it into your ERP, accounting and CRM platforms through their APIs. Four rounds of revisions are included.
- Working system tested before it touches live records
- API connections to platforms such as Xero, MYOB, HubSpot or Salesforce
- Progress updates through every phase
- [ Stage 04 ]Go-live
Production Handover & Training
Move the system into live operation and train your team on its real workflows and tools. You own everything engineered for your business, and support stays free after delivery until you choose a retainer.
- Production deployment and documentation
- Hands-on team training on the live system
- 60% payment due only after testing and acceptance
Commitments in writing
These terms are part of every engagement and are set out in the Kernel Flow terms and conditions.
- KPIs in writing
- Results are defined as measurable KPIs, signed by both parties after blueprint approval and before the build begins.
- Pay on approval and acceptance
- 40% after you approve the blueprint. 60% after the system is completed, tested and accepted by your team.
- 4 revision rounds
- Each build includes four rounds of feedback and changes. Anything outside the agreed scope is quoted before work starts.
- Free support after delivery
- Bug fixes, minor adjustments and troubleshooting are covered at no cost until you start a monthly retainer.
- Optional retainer
- From $500 per month for ongoing improvements, updates and priority support. Cancel with 30 days written notice.
- You own the system
- All workflows and systems engineered specifically for your business belong to you.
Security and data handling
Full detail is in the privacy policy.
- Credentials
- System credentials and API keys are stored encrypted, with access restricted to authorised Kernel Flow team members. They are deleted within 30 days of project completion, or immediately on request.
- Test data
- Builds are configured and tested on sample data, which can be anonymised or fully synthetic. Test data is never shared with third parties and is deleted within 30 days of completion.
- Data retention
- Client data is permanently deleted 30 days after the project or retainer ends.
- Privacy law
- Data handling follows the Privacy Act 1988 (Cth) and the Australian Privacy Principles. Client business data is never sold or traded.
- Stack
- Orchestration runs on n8n Enterprise, AWS Lambda and Vercel Edge. Language models come from OpenAI and Anthropic, connected to your platforms through their official APIs.
Get a Free AI Implementation Blueprint
Highest ROI Focus
Automate the key parts of your business that protect profit margins. Focus AI only where it delivers the highest return on investment.
Mapped Team-by-Team
Analyze your business departments to find exactly where AI can save time and speed up growth. Get a clear plan to remove manual roadblocks.
Direct Integrations
Connect AI systems directly into the tools your team already uses. Remove manual data entry and save hours of work every day.
Hover over the workstream cards or database item blocks to reveal specific AI implementation guidelines.