Roberto Pisati

A web application to digitalize and automate the car return process. At the end of each lease contract, cars are returned to an authorized LeasePlan dealer, which performs an extensive set of visual and technical inspections to assess the vehicle's condition. Results are recorded in a checklist, and pictures are taken. Damages beyond normal wear and tear are then invoiced to the customer.

The Inspection-to-Invoice application replaced the cumbersome manual process of filling a paper checklist, scanning it, sending it to the invoice department who would print it, manually inputting values, and invoicing customers with a modern and sleek solution, getting the job done with just a few clicks.

As a lead developer on this project, I designed the application architecture, developed the back end, worked closely with the front-end developer who designed the UI, and engaged with many teams in different parts of the world to assist them in connecting and implementing the new processes locally.

Accurate food intake information is important for a number of purposes, such as individual health and fitness, academic research, monitoring regulations, and commercial usage. Yet, it's so difficult to gather reliable data! We still rely on manual diaries or surveys, which are tedious and time-consuming.
Piccabi is a groundbreaking AI model that recognizes food and provides an accurate estimate of the portion size. All with a simple picture, taken with a mobile phone.

As a founding member, I've been part of an exciting melting pot, helping out wherever possible and getting stuck in technical and non-technical tasks. During this time, I've learned to deal with stakeholders at different levels and quickly adapted to the challenge of presenting and defending ideas in front of people whose job is to "pull you apart."
The project is on standby as we never managed to get sufficient funding or support, but it has nevertheless been an incredible adventure and experience.

One of the core processes in the Department of Social Protection is the nightly payments batch job suite, responsible for processing every single schema and allowance they support.
As part of a global modernization process, one of the projects I've been tasked with was to automate their legacy solution, based on an old .NET WinForms application where operators were still manually inputting parameters, values, and clicking buttons to kick off the job and review their status upon completion.
After extracting the core jobs functionality from the WinForms to a core set of libraries, I've been able to package them up into a command line tool (a console application) so that they could be executed via DOS commands or PowerShell. This put the Department in a position of going out for a tender for a generic scheduler tool and selecting Control-M (Java based) as their preferred choice.
Working closely with the operator's team, we automated the scheduling, execution, monitoring, incident creation, and reporting processes to the point that the requirement for people on a night shift went from eight persons to two, one actively working and one on-call.