Healthcare
Practice software that also satisfied the National Health Fund
A general physician practice needed its own management system. Day to day that meant appointments, medical history, and a patient registry. A lot of history still arrived on paper and had to be scanned in.
Nothing off the shelf existed for this at the time. Several doctors and nurses had to use the same system, with access control. Medical data had to be encrypted at rest. National Health Fund reporting meant complex statistics that were hard to produce by hand.
Designed and built a multi-user Windows client with a custom backend and database. It covered booking, rescheduling, and cancelling appointments; medical history notes, including paper scans and OCR; a managed patient registry; access control for doctors and nurses; encryption at rest; and the National Health Fund reporting.
One system covered the practice's automation: appointments, records, the patient registry, and fund reporting. Complex statistics for the National Health Fund were no longer troublesome to produce. Paper history could be stored and searched instead of sitting in a file.
Banking
Treasury trading at a fraction of the monopoly-platform cost
Treasury departments in banking communities needed to communicate and trade foreign exchange, fixed income, and money markets. The communities were national, industry-wide, and international.
Messages had to stay in time order. The architecture had to be distributed. Communication and trading had to be secure and auditable for treasury desks. Large monopoly platforms offered the function, but at a high price and with poor fit for each bank's own back office.
Key part of the design, architecture, and development of an international FX, fixed income, and money-markets system for banking communities, national, industry-wide, and international. The platform kept message order in time, ran as a distributed system, and gave treasury departments secure, auditable communication and trading. Integrated it with back-office systems so trading tickets could be automated, including bespoke back offices.
Banks got similar value to the large monopolized platforms at a fraction of the cost, plus straightforward integration with the back-office systems they already ran.
Transportation
Brokerage software for loads that don't fit the standard
A startup wanted to specialise in brokerage for oversized, non-standard, and ADR (dangerous goods) loads. They needed software that matched their own operational workflow, not a generic freight package.
Non-standard transport is hard to price quickly. Client, carrier, and driver all had to see load position, status, and timetable changes. Profitability of the brokerage had to be visible. Load details often arrived as free text and had to become strongly typed attributes. There was no AI to do that conversion at the time.
Designed and built the system. A calculation engine priced non-standard transport. The platform messaged both the client and the transportation entity about current load position and status. Drivers had thin clients on their phones for status reporting and timetable changes. The system tracked brokerage profitability and turned free-text descriptions into strongly typed attributes without AI.
They could process client requests efficiently and price non-standard loads quickly. Clients and carriers stayed informed on position and status. Drivers reported from the phone. The brokerage could see whether a job was worth taking, with structured data behind the messy descriptions.