AI engineer. Melbourne.
Not an agency. Not a team of consultants.
I build AI agents and MCP
infrastructure for founders, fast.
Claude APIMCP ProtocolLangChainCrewAI
n8n
ZapierWindmill
Next.js
Docker
Shopify PlusSelected work
$ storealert.monitor('store-events')
> Price change detected: SKU #4821
> Alert dispatched → 3 recipients
Real-time Shopify event monitoring: product changes, price updates, deletions. Live on the Shopify App Store.
$ eva.mcp.connect('warehouse')
> Inventory sync: 2,847 SKUs
> ERP bridge: ready
Eva MCP
MCP serverProduction MCP server bridging AI assistants to a proprietary mini ERP. Inventory, fulfillment, and warehouse management.
$ timer.start()
> 00:00 - scaffold
> 07:30 - shipped to prod
Supertrove
ProductFull product, concept to production. 7.5 hours.
$ ten4.voice.init('ndis')
> Provider connected
> Voice interface: active
Ten4AI
AI voiceAI voice assistant built for NDIS service providers in Australia. Handles provider queries, compliance workflows, and client communications.
Why not an agency
| Ali Kazim | Agency | In-house hire | |
|---|---|---|---|
| Time to start | This week | 2–4 weeks | 3–6 months |
| Who you work with | Me, directly | Account manager | Whoever you hire |
| Commitment | Flexible | Project retainer | Salary + benefits |
| Overhead | None | Agency margins | HR, onboarding, equity |
●Direct access to the builder. No middlemen, no overhead, no delays.
Skip the agency queue.
Start building this week.
Book a callAbout
I am an AI engineer based in Melbourne. I work with founders and operators across Australia and New Zealand to build AI infrastructure that ships: agents that automate real workflows, MCP servers that connect tools and data to any AI model, and full-stack products built in weeks rather than quarters.
Before going independent, I worked at enterprise scale on internal AI marketplace platforms, multi-tenancy SaaS infrastructure, and researched and introduced Agentic Software Engineering practices at organisations operating across Australia. That work required AI systems that ran reliably under real load, not just in demos.
What I do now is smaller in scope and faster to deliver. I take on a limited number of clients at a time, work directly with the person who owns the problem, and build rather than consult. Most engagements run four to eight weeks. The output is working software, not a strategy document.
The proof is in the shipped work.
Melbourne-based. Founder-first. Available for the right problem.