Advisory · Build · Enablement
Custom internal software and practical AI enablement, production-grade — usually on Lovable, always on the right platform for the job — in weeks, not quarters.
Too small for an engineering sprint. Too specific for any SaaS product. The ops team stitching together spreadsheets, Zapier and Slack. The finance team losing three days a month reconciling five systems. The report nobody has time to automate. For thirty years the answer was "wait for the roadmap" or "pay for software that's 70% right."
That maths has changed. When building costs a fraction of what it used to, the question stops being "can we afford to build this?" and becomes "why are we still paying for software that doesn't fit?"
We embed with your team, find the highest-value tooling and enablement gaps across your business, and build them — internal tools, dashboards, client portals, approval workflows, reporting, AI enablement — as real, loadbearing software your team relies on every day.
We build on Lovable where it fits, and wire into the tools you already run. You own the code, on your infrastructure. We keep it running and keep finding the next thing worth building.
Four steps. Transparent the whole way.
A short pre-work questionnaire and one working session. We map where time and money leak, and what you wish someone had built you years ago.
You get a costed, phased build scope — prioritised by value, no black boxes, no open-ended hours. You see exactly what you're buying before we build a thing.
We build in phases on Lovable: a working prototype first, then the tools around it, then the loadbearing software your business runs on. Production-grade — real data, authentication, security.
Ongoing support and iteration, and we keep surfacing the next tool worth building across your functions.
Strategic judgement on one side, AI-native delivery on the other — every Invictus engagement stands on both. That's the pairing traditional agencies can't match on speed, and pure-AI tools can't match on judgement.
Human Intelligence
The judgement that sets the direction.
AI Intelligence
The precision that executes it.
AI identifies the patterns. People determine the path forward.
The best outcomes aren't human or AI — they're both. Human judgement finds the real problem and sets the direction; AI brings the speed, scale and precision to execute it, and to see what a team alone would miss.
“AI identifies the patterns. People determine the path forward.”
We run our own business on Lovable, Claude and agentic workflows — the same stack we build yours on. That's conviction you can't fake, and speed a traditional agency can't match.
Development is commoditised now; the value is everything around the build — finding the bottleneck, articulating what it's worth, and shipping the thing. We do all three.
We start from your problem, not a product. Often that means building on Lovable — our core engine — and wiring it into the tools you already run, like Google Workspace, monday.com and Claude. Enterprise-grade foundations, chosen to fit.
The software we build for you is yours — the code, the logic, the data. Where a build runs on a platform like Lovable or monday.com, you hold the subscription directly, so you're never locked to us as a middleman.
Delivering sustainable commercial performance through Human Judgement + AI Capability.
Invictus Management Consulting is led by Craig Zeelie. For more than two decades I led businesses through the moments that decide their future — turnarounds, transformations and growth resets — in senior roles from General Manager to COO and Managing Director, across Australia, Africa and Asia Pacific, and across technology, manufacturing, FMCG, retail services and global B2B. I've carried multi-country P&Ls, rebuilt margins, and run full turnarounds. That's taught me exactly where time and money leak inside a business — and how rarely the right tool exists to stop it.
I founded Invictus on a simple conviction: the cost of building software has collapsed, and most businesses are still paying for tools that are only 70% right. So I build the ones that fit — on a Dual-Pillar model I practise, not just preach: human judgement to decide what's worth doing, AI to build it with speed and precision. Invictus runs on Lovable, Claude and agentic workflows — the same stack I build for clients — and I stay hands-on in the build, not just the brief.
Invictus means "unconquered" — it's how I lead, how I consult, and how I help organisations rise: stronger, sharper, and ready for what comes next. If your business is sitting on a problem worth solving, I'd welcome the conversation.
Selected experience: Avery Dennison (GM, Sub-Saharan Africa) · Weetabix (GM/MD) · Bytes Document Solutions (Divisional MD) · Smollan (10+ years, GM/Divisional/Regional) · SPARFACTS Australia (COO).
Education: Stellenbosch University · GIBS Business School.
A sample of work — the problem, and what changed.
A 54-second look at the field-data build — ten spreadsheets to one live decision tool.
Operations re-engineering — five systems and a hand-typed key, rebuilt into one source of truth. Read the full case study →
A portable dashboard, not a locked-in report
What it was: Seven rounds of field submissions and ~1,400 free-text rep comments arrived as ten disconnected spreadsheets — anecdotes, no defensible view, and nothing shareable.
Outcome: A live dashboard built on Lovable that names the gap stores, prices the risk and recommends the call-cycle change — with one-click Excel export and a standalone HTML file you can email to anyone, filters intact, no login. The data stays yours.
Five drifting systems, re-joined by machine
What it was: A field operation ran across five disconnected systems joined by one hand-typed key; when the ops manager left, accuracy fell to 42% and invoices were being missed.
Outcome: Confirmation now triggers the pipeline automatically — execution board, invoicing and a fixed-cadence chase — so directors see the backlog every Monday instead of discovering a cash-flow gap late.
A hiring pipeline that runs itself
What it was: Recruitment, onboarding and training spanned seven monday.com boards and depended on people remembering to send emails and mark results by hand.
Outcome: One status change routes a candidate through quizzes, live-session attendance, grading and a hard compliance gate — unattended.
A live field-ops command centre
What it was: A field-execution operation had no single view of what was staffed, completed or slipping — visit performance was pieced together after the fact.
Outcome: A live Looker Studio dashboard tracking every visit — total, unstaffed, completed, due today and pace-to-target — with missed and pending visits by program, and filters so each regional lead sees only their territory. The view a director reads in ninety seconds.
Operations, finance, HR and revenue leaders sitting on a backlog of "someone should build us a tool for that." Companies paying for a stack of SaaS that's each 70% right. Teams told "the roadmap is full" one too many times.
Placeholder copy. A 30-minute conversation, no obligation, and a clear next step either way.