Axel AI is a one-person AI automation studio operating from Brisbane, Queensland. I build custom workflow automation, backend pipelines, reporting systems, and internal tools for SMB teams across Australia and (remotely) the rest of the world.
The whole operation runs on a single principle: the person who scopes the work is the person who writes the code. No account managers, no sales layer, no offshore handoff. You talk to the engineer, every meeting, every revision.
What "custom AI automation" actually means here
Most "AI" products on the market are chatbots — you ask, they answer. The systems I build are infrastructure: long-running pipelines that take action, ask when uncertain, and finish the work without prompting. Invoice routing. Vendor reconciliation. Weekly report assembly. CRM lead enrichment. Internal tools matched to your team's actual workflow. Each one runs on deterministic scripts where execution matters, with AI used only for the reasoning steps that need it.
How engagements work
- Audit(1–2 weeks, A$2–5k) — embedded with your team, mapping every workflow, instrumenting the time-sinks, writing up what's worth building and why.
- Build (typically 2–5 weeks, fixed scope) — weekly demos in your own Slack, every workflow living in your repo from day one.
- Hand-off & optional retainer — code in your repo, automation on your infra, runbooks in your wiki. Two of your engineers trained. Retainer is monthly, cancel-anytime.
Operating principles
- You own everything. Code in your repository, configs on your accounts, infra you control. No platform lock-in, no recurring licence fees.
- Boring tech where it works.AI where reasoning matters, plain code where it doesn't. Your token bill should reflect that.
- Honest scope.If a workflow shouldn't be automated, you'll be told. I'd rather lose the build than ship something brittle.
- Demo every seven days.No big reveals. You see the work as it's made, in your own Slack, every week.
Stack I work in
Python, TypeScript, Postgres, Snowflake, Modal, Temporal, Slack, Linear, Notion, Zendesk, Salesforce, Hubspot, Quickbooks, Stripe. Other stacks happily — these are just the ones I reach for first.
The founder
Nile Bugler — Founder & AI Automation Engineer.
Axel AI came out of running a service-based business and watching how much of every day vanished into admin — invoicing, ads reporting, follow-ups, the steady drip of inbox triage. Rather than outsource it or hire for it, Nile built the automations himself: an email scanner that triaged client mail by intent, an automated accountant for receipts and reconciliations, an ads-reporting pipeline, and a personal-assistant agent that handled the work between the work.
The result was hours reclaimed every week — and a clear sense of which problems AI actually solved and which ones it didn't. Two years building production software taught Nile where reasoning matters and where plain code wins.
Axel AI exists to bring that same engineering judgement to other small teams that can't justify a full-time automation hire — and to do it without an agency layer in between. The person who scopes the work is the person who writes the code.
Where to next
The fastest way to find out whether your workflows are a fit is the audit week. Send me a message with the workflow that's eating the most hours. I'll reply within 24 hours, Brisbane time, M–F.