Inspiration

Part of the inspiration comes from having to wear so many hats during the pandemic. To having to coordinate schedules across many different projects, from developing medical devices (current job) to scheduling kids getting schooled right at home over two households. Monday brings simplicity to the task of scheduling and tracking a team. But when the tasks get small (assignments) and there are many people with many dependencies, creating a schedule is problematic. ByNextMonday aims to put that load onto the computer and synchronize a very visual board with the Monday database. By letting the computer solve the most efficient routes, I think it could be bring monday to new areas, like education and even personal health care (scheduling at home visits ). And, simplifying the generation of scheduling increases the amount that can be scheduled.

What it does

BNM reads from a stream of jobs (collections of tasks) and users, transforms them (say, a school lesson to lessons for every student), and distributes them over user's available time, casting the right tasks to the right actors and optimizing all of this with constraints of time, location, skill, etc.

Currently, it has to be hardwired to read jobs from /assets/classroom_jobs.json /assets/classroom_users.json

How we built it

The front end is an angular web app. The backend is python on top of mathematical libraries. Firebase handles the primary storage and communications issues

Challenges we ran into

Currently, we can solve the schedule more efficiently that we can scale number of tasks written into the monday db, There definitely may be a better way to do this and I'd love to work with Monday to figure this out. Unfortunately, the db calls to create items had to be throttled down to 1 item creation per 4 seconds or the server would start throwing errors and all the items could not be created. I'd love to find out how to programmatically batch import into the system.