Meridian Labs

Selected work

Judged by what's running in production.

A selection of engagements across software, AI and automation. Each follows the same structure: the problem, what we built, and what changed.

[ project visual placeholder ]

Custom Software · Automation

Operations Platform for an NDIS Provider

The problem

A growing provider was managing 200+ participants across spreadsheets, a generic CRM and email. Claims were delayed, documentation was inconsistent, and audit preparation consumed weeks of staff time.

What we built

We built a purpose-designed platform covering referrals, service agreements, rostering, progress notes and NDIA claiming, with an AI drafting layer for notes that staff review and approve.

Outcomes

  • Placeholder metric — e.g. claiming cycle reduced from 14 days to 2
  • Placeholder metric — e.g. 30+ admin hours saved weekly
  • Placeholder metric — e.g. audit preparation time cut by 80%

Stack

Next.jsPostgreSQLClaudeAWS
[ project visual placeholder ]

Automation · Integrations

Enrolment Automation for an RTO

The problem

Every enrolment was typed into three systems by hand. Errors crept into compliance records and follow-ups were missed during intake peaks.

What we built

An automated pipeline validates applications, generates compliant documentation, syncs the student management system and schedules follow-ups — with exceptions routed to staff.

Outcomes

  • Placeholder metric — e.g. enrolment processing time down 90%
  • Placeholder metric — e.g. zero data re-entry across systems
  • Placeholder metric — e.g. enquiry response time under 5 minutes

Stack

Node.jsn8nREST APIsAzure
[ project visual placeholder ]

AI Agents

AI Documentation Assistant for Allied Health

The problem

Clinicians were spending 90+ minutes a day on documentation after sessions, delaying reports and driving burnout.

What we built

A secure drafting assistant grounded in the practice's templates and policies produces structured notes and report drafts for clinician review, then files approved documents to the practice system.

Outcomes

  • Placeholder metric — e.g. documentation time reduced by 65%
  • Placeholder metric — e.g. report turnaround from 3 weeks to 4 days
  • Placeholder metric — e.g. 100% of drafts human-reviewed before filing

Stack

ClaudePythonFastAPIAustralian-region hosting
[ project visual placeholder ]

Enterprise · Cloud

Integration Layer for a National Services Group

The problem

Four acquired businesses ran four incompatible systems. Consolidated reporting took finance two weeks each month.

What we built

A central integration layer with a shared data model streams events between systems in near real time, feeding a live executive dashboard.

Outcomes

  • Placeholder metric — e.g. month-end reporting from 2 weeks to 1 day
  • Placeholder metric — e.g. single source of truth across 4 entities
  • Placeholder metric — e.g. 99.95% pipeline uptime

Stack

AWSTypeScriptPostgreSQLGraphQL

A note on confidentiality

Much of our best work is under NDA.

The case studies above are representative placeholders. In a discovery call we can walk you through detailed, relevant examples — including systems similar to what you're planning.

Want a system like these?

Bring us the operational problem. We'll bring the architecture, the estimate and the delivery plan.

No obligation · Fixed-scope proposals · Reply within 1 business day