Case Studies
A selection of delivered client projects and capability showcases that demonstrate the range of work we do.
Client Projects
Paid engagements delivered to real clients.
REMIRA Intelligent Software Solutions
Global POS & ERP Support
Senior Consultant supporting 1,500+ POS devices and enterprise retail systems across Australasia, Southeast Asia, Africa, Europe, and South America.
Challenge
REMIRA's retail clients required seamless support across multiple time zones and regulatory environments, including mandatory fiscalisation and e-invoicing compliance in 11 countries.
Solution
Provided end-to-end consulting across server maintenance, database management, custom integrations, API testing, and project management. Led multi-country fiscalisation rollouts and coordinated cross-timezone upgrade projects.
Key Results
- Remote upgrade of 400 POS terminals and Head Office server in the US within 12 hours
- Fiscalisation delivery across Germany, Spain, Italy, Poland, Hungary, Austria, Colombia, Peru, Ecuador, Kenya, and Cyprus
- 2-stage POS upgrade of all Politix Australia terminals with zero downtime during trade
- Built AI MCP server for natural language ERP report interpretation
- Developed Python receipt backup solution delivered as Windows executable
Legacy Brazilian Jiu Jitsu
E-Commerce & Reorder Intelligence
Built Shopify e-commerce store and SaaS reorder analysis tool for a $200K fight wear and uniform merchandise operation.
Challenge
Legacy BJJ needed an online sales channel and a way to manage reorder decisions. The existing process involved 3-hour manual procedures with Shopify report exports and Excel lookups.
Solution
Created the Shopify e-commerce store with wholesale pricing for franchise branches. Built a SaaS reorder analysis tool on Vercel with Neon Postgres that reduced the reforecast process from 3 hours to seconds.
Key Results
- Shopify store launched with wholesale pricing model for franchise branches
- Reforecast process reduced from 3 hours to seconds
- Fully SaaS-based solution on Vercel with serverless Postgres
Berry Brazil Acai
Sales Forecasting & Invoicing Platform
SaaS forecasting and invoicing platform for a 104-ton annual acai distribution operation.
Challenge
Berry Brazil managed stock forecasting via spreadsheets and produced invoices manually in Word documents saved to Google Drive. The delivery team had no mobile-friendly tools.
Solution
Built an online SKU forecasting tool that scrapes invoices from Google Drive, populates historical sales data, and enables re-forecasting. Developing a mobile invoicing app for field delivery with automated email delivery via Resend.
Key Results
- Automated invoice scraping from Google Drive
- Historical sales comparison against targets for reorder decisions
- Mobile-optimised interface for warehouse and field sales staff
- Migration from AWS to Vercel + Neon Postgres for simpler deployment
Showcases
Capability demonstrations built to showcase what we can deliver.
Square Stock Control
Multi-Tenant Reorder Intelligence
Standalone multi-tenant app that syncs Square POS data and produces deterministic reorder recommendations at variant and item level.
Challenge
Small retailers using Square POS lack built-in reorder intelligence. They need automated stock velocity calculations and suggested reorder quantities without expensive enterprise tools.
Solution
Designed a Next.js application that ingests Square Catalog, Inventory, and Orders data via API. Computes weekly sales velocity, weeks of cover, and suggested reorder quantities with parent-item rollup views.
Key Results
- 8-week sales velocity calculation with configurable lead times
- Variant-level and parent-item rollup views with expand/collapse UX
- Square OAuth integration with webhook-based incremental updates
- Multi-tenant architecture ready for multiple retail clients
AI & MCP Development
AI-Powered ERP Reporting
MCP server that enables natural language querying of ERP reports, turning complex retail data into human-readable insights.
Challenge
ERP systems produce data-heavy reports that require specialist knowledge to interpret. Business users need insights without learning SQL or navigating complex report interfaces.
Solution
Developed an MCP server that exports standard ERP reports into a structured database, columnises the data for efficient querying, and trains a language model on domain-specific terminology to produce accurate, natural language insights.
Key Results
- Natural language interface for ERP data queries
- Domain-specific terminology training for retail metrics
- Structured database export from standard ERP reports
- Accurate insights without SQL knowledge required