shazil_mohammed / melbourne

Shazil
Mohammed

I translate business problems into AI workflows,
then ship them.

Melbourne, Australia
Open to remote, hybrid, on-site
Available now
$capabilities.sh --list
Workflow automationAPI integrationPrompt engineeringAgent deploymentStack integrationProcess automationMulti-agent orchestrationRAG architectureToken cost optimisationLead-gen automationFunnel designBusiness translation
$projects.sh --selected
[ 01 ]
9-to-5 Exit Blueprint Live
Next.js 14TypeScriptAnthropic APISupabaseConvertKitGumroadVercel
Problem
Build a self-sustaining digital product business converting cold social traffic into paid customers — no team, no ad spend.
What I built
Five-page animated quiz generating personalised result profiles via streaming Anthropic API calls. System prompt caching and compact input formatting bring cost-per-completion under one cent. 29-page PDF product, three-email ConvertKit automation, Gumroad delivery, full security middleware and rate limiting on Vercel.
Why it matters
End-to-end product shipped solo. AI is the conversion engine, not a feature. Token caching, batching, JSON-only outputs, and session deduplication architected in from day one.
exitblueprint.com.au — Side Income Quiz
Question 1 of 3
What's your current situation?
Your Side Income Profile
[ 02 ]
CoolFlowBuilt
Node.jsFirebaseAnthropic APIGoogle Calendar APITwilioNodemailerScrapingBee
Problem
A Queensland HVAC contractor needed a sales pipeline that ran without a sales team — daily lead flow, qualification, outreach, and booked callbacks.
What I built
Full multi-agent operations platform. Daily AEST-scheduled scraping of Google Maps, Airtasker, hipages, and council DA feeds. AI-driven BANT lead scoring, automated email and SMS outreach, Google Calendar-confirmed callback booking. Five-page CRM: dashboard, leads, kanban, bookings, and Carrier/Toshiba unit catalogue.
Why it matters
Take "fix our sales process" and ship the entire stack, from scraping rails to outreach cadence to the CRM the team operates from.
CoolFlow · QLD Lead Pipeline · Today 07:00 AEST
47
Scraped
23
Qualified
8
Booked
BusinessCategoryScoreStatus
Brisbane Air ConditioningHVAC Contractor87 HOTEmail + SMS + Calendar ✓
Gold Coast Heat PumpsHVAC Contractor74 WARMEmail sent
Coastal Climate ControlRefrigeration62 WARMEmail sent
Metro Plumbing & GasPlumbing38 COLDFiltered out
[ 03 ]
Melbourne Brand OSBuilt
Anthropic APINotionMake.comBufferFigmaWebflow
Problem
A creative director with 25 years of experience needed to scale beyond billable hours without losing the senior judgement that justifies premium pricing.
What I built
Nine-section operating strategy and a seven-stage AI workflow — brand intake, brand brain, content generation, calendar, approval gate, distribution, optimisation. Three productised service tiers. Year-one revenue projection $145k–$250k recurring. Full Melbourne go-to-market plan.
Why it matters
AI positioned to augment senior human judgement — not replace it. Service-first economics that evolve into SaaS.
Melbourne Brand OS — AI Workflow
[ 04 ]
Hydraulics Engineering AgentArchitected
Claude APIRAGAS/NZS 3500Custom calc engineCAD/BIM APIs
Problem
Senior hydraulics engineers spend hours cross-referencing standards and running calculations — work an agent could surface in minutes if accuracy and traceability were built in correctly.
What I built
Orchestrator routing to four specialist sub-agents — design calculations, compliance, field diagnostics, documentation. Three-block system prompt strategy: persona/scope, chain-of-thought with explicit assumptions, structured output enforcing summary + calculations + standards citations + risk flags. RAG over actual AS/NZS 3500 documents.
Why it matters
Architected for trust. Aggressive uncertainty flagging built into every prompt. Agent positioned as reviewer above human judgement — the only viable shape for compliance-sensitive AI.
Hydraulics Agent — Multi-Agent Router
ORCHESTRATOR — classifying query → routing to specialist sub-agent  awaiting query
Design
Calculations
Compliance
Review
Field
Diagnostics
Documentation
Writer
Summary
Calculation
Citation
Risk flag
Confidence
[ 05 ]
Soul MirrorBuilt
HTML / CSS / JSAnthropic APICanvas APIPDF generation
Problem
Build a consumer identity product that feels personal, mystical, and shareable — without an app store cycle or a real backend.
What I built
Gamified self-discovery app combining MBTI, astrology, numerology, and love languages into a unified Soul Profile. Pixel-art avatars, 8 characters × 4 emotion states with evolution unlocks, AI-generated narrative reports, PDF download, profile comparison, 2D avatar chatroom — all in a single HTML file (~7,700 lines).
Why it matters
Proves capacity to finish complex consumer-facing work end-to-end. Design, UX, AI integration, rendering engine, and ship — without losing aesthetic coherence under scope pressure.
Soul Mirror — Choose Your Archetype
The Architect
The Guardian
The Seeker
The Creator
[ 06 ]
Cowork × Claude Code OrchestrationLive Workflow
CoworkClaude CodeVercel CLIMarkdown handoff protocol
Problem
Multi-platform setup tasks require both browser interaction and terminal commands. Normally that's two separate human passes.
What I built
Paired prompt files that hand off cleanly between two agents. Cowork drives the browser — creates accounts, builds automations, gathers credentials. Claude Code injects credentials, deploys via Vercel CLI, runs validation tests, writes a launch summary. One human kick-off. Two agents finish the work.
Why it matters
Most operators are still choosing between single tools; this is two specialised agents working a shared deliverable through a defined handoff protocol.
Agent Orchestration — Live Workflow Run
Cowork / Browser Agent
Claude Code / Terminal
$principles.sh --list
01.
Ship over architect-in-slides
Bias toward a deployable pilot in week one over a polished plan in month one. A working prototype surfaces real constraints that no slide deck ever will.
02.
AI is a workflow component, not a feature
The model sits inside a larger system alongside scraping, scheduling, payments, and human review. The design problem is the workflow around the call — not the call itself.
03.
Unit economics from day one
Prompt caching, compact inputs, JSON-only outputs, batching, session deduplication, fallback paths. Cost control is a prompt-level design constraint — not a billing problem you solve later.
04.
Tight loops, measurable steps
Pilot week one. Measure week two. Scale or kill week three. If you can't define what success looks like before you build, you're not ready to build.
05.
Trust before autonomy
In compliance-sensitive verticals, architect agents as reviewers above human judgement before promoting to autonomous operators. Uncertainty flagging built into the prompt — not bolted on after.
$contact.sh
Location
Melbourne, Australia
Availability
Open to remote, hybrid, or Melbourne-based