Custom platforms in production for our clients.

Custom Laravel platforms for operations where off-the-shelf SaaS won't fit — mining dispatch, multi-tenant SaaS, digital business cards at corporate scale. Four named platforms still running: Palabora Mining Company, Cullinan Diamond Mine, DM3, BusinessLink. Project-quoted by scope.

Talk to us

What custom platforms actually do

A custom platform is what you build when off-the-shelf SaaS imposes a workflow that doesn't fit the operation. The shape is always the same: a Laravel application stitched into the systems and hardware that were there before you arrived, with domain logic the off-the-shelf product doesn't model — block-cave draw control, dispatch shift handovers, per-tenant data isolation enforced at the route layer, batch onboarding for hundreds of users in one CSV. The unglamorous parts are what make the platform credible — audit trails the inspectorate actually reads, dual-write paths so a network drop mid-shift doesn't lose production data, MQTT brokers that survive patchy underground WiFi, role hierarchies that respect three different ways the same company manages access. Four named platforms still in production — two in mining, one as the agency's own monitoring SaaS, one for an enterprise corporate group — covers most of the shapes the work takes.

Engagements

Palabora Mining Company

Real-time dispatch and production management for Palabora Mining Company's block-cave operation. Laravel + PostgreSQL + MQTT over SSL, 24/7 coverage across six user roles.

Cullinan Diamond Mine

Underground LHD dispatch system at Cullinan Diamond Mine. Ten-plus years in production, four active mining levels, 24/7/365 coverage. Laravel + Livewire + SQL Server.

DM3

Strata Logic's own SaaS monitoring platform. Aggregates 10+ external APIs into a unified dashboard for every client website; live since February 2024. Laravel 11 + Livewire 3 + PostgreSQL.

BusinessLink

Multi-company digital business card platform for enterprise corporate groups. QR + NFC delivery; batch CSV import for hundreds of employees; Laravel 11 + Livewire 3.

How we approach custom platforms

  • 1. Discovery

    Two to four weeks. Stakeholder workshops with operations, IT, and the people who'll actually use the platform every day. Audit of existing systems the new platform will integrate with — ERP, legacy databases, third-party APIs, file-based imports from supplier systems. The deliverable is a written scope with a stack named against it.

  • 2. Architecture

    Data flow sketched end to end before any code. Tenancy model decided up front (single-tenant, multi-tenant per company, multi-tenant per user), with role hierarchy specified at the same time. Integration boundaries named — what's a real-time webhook, what's a nightly batch, what's a one-way feed. Redundancy posture for the connections you can't control.

  • 3. Integration

    ERP endpoints, third-party SaaS APIs, file-drop FTP from suppliers. We adapt to whatever's there — most platforms are joining a stack that's been running for years, not greenfield. Each integration gets its own resilience posture: what happens when the upstream API is down, what happens when the file is malformed, what happens when authentication tokens expire.

  • 4. Handover

    Training is per-role, not per-user. Admins, managers, and operators all learn the platform differently and need different runbooks. IT team gets a written deploy/rollback document and on-call escalation path. Cutover is scheduled against operational reality — never mid-shift, never mid-month-end, never the day before a board meeting.

  • 5. Maintenance

    A separate enterprise retainer, not a Site Care plan — scale and SLA are different. Quarterly review covering incidents, capacity, and roadmap. Named on-call engineers, patch cycles scheduled around the operation rather than vendor-default Patch Tuesday. Cullinan Diamond Mine has been running this way for the better part of a decade.

Common questions about custom platforms

What counts as a custom platform?
Anything where the off-the-shelf product won't carry the workflow. Block-cave draw control wasn't in any vendor's catalogue when Palabora Mining Company needed it. Multi-company digital business cards with per-company branding and batch CSV import weren't in any digital-card SaaS we could find. The agency's own monitoring platform (DM3) replaced roughly R10-50k/month in third-party tools per client. The category isn't "we needed something cool" — it's "the catalogue doesn't fit the operation".
Why Laravel + Livewire?
Default stack across every platform we run: Laravel for the backend, Livewire 3 for the reactive UI without a separate JavaScript front-end, PostgreSQL or SQL Server depending on the operational reality (most mining sites have SQL Server, most SaaS prefers Postgres). Spatie Permission for role-based access control on every platform. The stack stays stable across engagements so the team that built it can maintain it without context-switching every project.
How long do these projects take?
6 to 18 months from discovery to handover, depending on integration depth and the number of upstream systems involved. Cullinan Diamond Mine was on the longer end (~18 months) given the underground network and custom HMI work. A multi-tenant SaaS rebuild on a stable stack lands closer to 6 months. We band-quote during scoping, not after.
Do you keep maintaining it after handover?
Yes — separate enterprise retainer, not a Site Care plan. Several clients have run with us 5 to 10 years past initial cutover. The team that built the platform is the team that maintains it; no developer-rota churn. Cullinan Diamond Mine has been running with us roughly a decade, Palabora Mining Company across multiple production years.
What about integrations with existing systems?
Standard part of every engagement — most custom platforms are joining a stack that's been running for years. ERP integrations, file-drop FTP from suppliers, third-party SaaS APIs. Each integration gets a resilience posture documented during architecture: what happens when the upstream is unreachable, how we authenticate, what the retry shape looks like.
Is the source code yours or ours?
Yours. Every platform we build is owned outright by the client at handover. We retain the right to reference the engagement (with the client's permission — the four published case studies all happened with explicit consent) but no proprietary lock-in, no per-seat licensing, no rented IP.

Custom Laravel platforms since 2013. Four named systems still running: Cullinan Diamond Mine underground for the better part of a decade, Palabora Mining Company at block-cave scale, DM3 as the agency's own monitoring SaaS since February 2024, BusinessLink in production for an enterprise corporate group across multiple companies. We maintain everything we ship — the team that built the platform is the team you call when something needs changing five years later. We think each client's problem through and build exactly what it needs — and we've done that, deliberately, for over a decade.

Smaller scope? See custom web development →

Got a platform to scope?

Talk to us