What is AI Strategy & Training?
AI Strategy & Training AI Strategy and Training is a structured advisory engagement that turns scattered AI use into a plan a team adopts. UnderCurrent Automations delivers it from Melbourne to Australian small businesses: current-state assessment, 90-day roadmap, vendor-neutral tool picks across ChatGPT, Claude and Copilot, role-based workshops, and three months of office hours.
It is not a single tool you buy; it is the orchestration layer on top of the tools you already pay for, tuned to how AI Strategy & Training runs in your business day to day.
What does the build include?
A 90-day AI roadmap the team uses, with prompts and policies they keep. 8 deliverables by default, each tuned to how your business already operates.
- 01Current-state assessment: tool stack, data readiness, policy gaps, team AI literacy
- 0290-day AI roadmap with use cases ranked by ROI, effort and data readiness
- 03Vendor-neutral tool selection across ChatGPT, Claude, Microsoft Copilot, GitHub Copilot, Cursor, Notion AI, Gemini and Perplexity
- 04Policy and governance framework: data handling, model access, prompt logging, Australian Privacy Act coverage
- 05Role-specific team workshops for sales, ops, finance, marketing and leadership
- 06Prompt libraries and playbooks the team writes in the workshop and keeps using
- 07One-on-one executive coaching for founders and leadership teams
- 08Three months of monthly office hours after handover, for questions, drift correction and new use cases
Who is it built for?
If the team is already using ChatGPT or Claude on the side, but nothing has been adopted at the business level and tool spend is going into seats nobody uses, this is the discipline we run before any automation work begins. Common across tradies, business services, consultants and allied health practitioners in Australia.
Tradies and field-services teams
AI training that reaches staff who are not at a desk.
Plumbers, electricians, builders, logistics crews, field techs. ServiceM8 or Jobber runs the jobs, Xero sends the invoices, and the owner answers quotes from the ute. We train field and office staff on voice-based ChatGPT and Claude for quote drafting, job-note clean-up and inbound-enquiry triage, with a short AI policy written around the tools you already pay for.
Business services, agencies and consultants
AI adoption that moves past the pilot and into billable work.
Accounting firms, agencies, strategy consultants, RevOps teams. HubSpot, Salesforce, Xero, Notion, Slack. We run role-based workshops for account managers, analysts, BDRs and partners on ChatGPT, Claude, Copilot, Cursor and Notion AI, then layer a policy draft so client-data exposure is controlled before anyone ships work to a public model.
Allied health practitioners
AI for the front desk and the clinical admin.
Physios, psychologists, osteopaths, podiatrists, GPs. Cliniko, Halaxy or Nookal runs bookings, HICAPS handles claiming. We train front-desk and practitioner staff on AI-assisted inbound replies, intake summarisation and referral drafting with ChatGPT and Claude, written to fit the Australian Privacy Act, health-sector standards and the practice management tool you already run.
What tools does it work with?
We build on the stack you already run. If a tool has an API, we can wire into it. Common integrations we deploy for AI Strategy & Training:
If you don't have a system in place yet, we help you choose a lightweight option that fits your business without adding overhead.
How does UnderCurrent build it?
Assess, plan, train, stay for three months. We move in weeks, not months.
- 01
Map
Audit how this works today. Sources, tools, handoffs, drop-off points. Output: a map with priority fixes.
- 02
Build
Deploy automation against your existing stack. Configured, tested, and documented end to end.
- 03
Flow
Hand over, train, monitor for 30 days. Results visible within 2–3 weeks of live.
How does it compare to alternatives?
Most Australian small businesses have sat through a ChatGPT Enterprise pitch, a Microsoft Copilot demo and a Big-4 consultancy one-pager. Nothing changed afterward. Rows are capabilities; columns are approaches. A check means the approach consistently delivers.
| Capability | UnderCurrent | Figuring It Out Internally | Vendor Training | Big-4 Consultancies |
|---|---|---|---|---|
| Vendor-neutral tool recommendations | ✓ | · | · | · |
| Role-specific workshops on real data | ✓ | · | · | ✓ |
| Policy and governance framework included | ✓ | · | · | ✓ |
| Delivered in weeks on a fixed timeline | ✓ | ✓ | ✓ | · |
| Post-engagement office hours for 3 months | ✓ | · | · | · |
| Small-business engagement pricing | ✓ | ✓ | ✓ | · |
What does it cost?
Project-priced, not per-seat. No ongoing software licence fees, because we build on the stack you already pay for.
from AUD $2,500
Workshop
Single session
A half-day role-based workshop for a team of up to 15. Prompt library, playbook and session recording included. Good entry point before a full engagement.
from AUD $5,000
Roadmap
Assessment + plan
Full assessment, 90-day roadmap, vendor-neutral tool selection, two role-based workshops and a written AI policy draft. The standard small-business engagement.
Contact for quote
Program
Team-wide
Roadmap plus role-based workshops for up to four teams, executive coaching, and three months of monthly office hours. Fits 20 to 100-person businesses.
How long does it take?
Most builds go live within 7 to 14 days of the first scoping call. Week one is discovery; week two is build and rollout. Measurable results are visible within 2 to 3 weeks of going live.
Full, multi-channel systems run 3 to 4 weeks end-to-end. We give you a firm timeline before committing, with no scope creep.
Where does UnderCurrent operate?
Based in Melbourne, Australia. We serve clients across Melbourne, Sydney, Brisbane, Perth, Adelaide, Canberra, Victoria, New South Wales, Queensland, Western Australia, South Australia, Australian Capital Territory, Australia. Most discovery and build work runs remotely; onsite sessions are available across greater Melbourne.
Frequently asked questions
How to use AI to improve a small business?
Start with three tasks you do every week that are slow, repeatable and low-risk: for example drafting quotes, writing follow-up emails, cleaning up job notes or researching prospects. Run a two-week pilot on ChatGPT or Claude with clear success criteria, then expand to the tools that earn their seat. UnderCurrent Automations runs this as a structured AI Strategy and Training engagement in Australia, covering assessment, a 90-day roadmap, vendor-neutral tool selection across ChatGPT, Claude, Copilot, Cursor and Notion AI, and hands-on workshops so the team adopts the work rather than bookmarks it.
How do small business owners use AI?
Owners of Australian small businesses use AI for three recurring jobs: drafting client-facing writing (quotes, proposals, follow-ups, social posts), making sense of incoming information (long emails, meeting notes, PDFs, spreadsheets) and automating small decisions (lead triage, invoice chasing, review replies). The common stack is ChatGPT or Claude for writing, Microsoft Copilot or Gemini in Office or Google Workspace for documents, and n8n or Zapier wired to Xero, HubSpot, ServiceM8 or Cliniko for the automated parts.
What is AI training for employees?
AI training for employees is role-specific, hands-on work that teaches staff how to use tools like ChatGPT, Claude, Microsoft Copilot and Notion AI on the jobs they do every day, with the data they handle every day, under a written policy that defines what can go into a public model and what cannot. UnderCurrent Automations delivers AI training for employees as half-day or full-day workshops for sales, ops, finance, marketing and leadership teams across Australia, producing a prompt library and a policy draft the team keeps using after the session.
How to teach employees how to use AI?
Run structured training in three stages, not a single all-hands demo. First, general AI literacy so the team understands how ChatGPT, Claude, Gemini and Copilot differ and what each is good at. Second, role-specific workshops where each team builds a prompt library on real tasks (a BDR on outreach, an accountant on reconciliations, a physio on intake notes). Third, a written AI policy covering data handling and approved tools, reviewed monthly. UnderCurrent Automations packages this as a three-stage training program for Australian small businesses, with three months of monthly office hours to correct drift and answer questions as usage scales.
What type of AI is good for business?
For Australian small businesses the useful categories are three. First, general-purpose assistants (ChatGPT, Claude, Gemini) for writing, research and structured-data tasks. Second, productivity-embedded AI (Microsoft Copilot in Office 365, Gemini in Google Workspace, Notion AI) for in-context drafting and review. Third, workflow and agent AI (OpenAI and Anthropic APIs wired through n8n, Zapier or Make) for actions that run without a human triggering every step. We pick the mix per business based on stack, budget and team literacy, and document the decision so nothing is locked to one vendor.
Why do 85% of AI projects fail?
The 85% figure, commonly cited from Gartner and RAND Corporation analyses, covers AI projects that never reach production or are abandoned within twelve months. The usual causes sit upstream of model capability: weak data readiness, unclear success metrics, no named internal owner, no change-management plan. UnderCurrent Automations addresses these with a current-state assessment, three to five measurable outcomes defined upfront, role-specific workshops that build owner habits, and three months of office hours so drift is caught before adoption dies.
Luke Marinovic
Founder of UnderCurrent Automations. Builds sales, content, finance and operations automation for Australian small businesses.
Related services and source material.
Adopt AI without burning a quarter on pilots that stall.
Book a free 30-minute call. We walk through what the team already uses, what is working and what is not, then sketch what a focused 90-day AI Strategy and Training engagement looks like for your business. No deck, no pitch.