I’m a former line cook who transitioned into tech, and I’m currently building a project called MEP (short for mise en place) with a scheduling frontend named Flo. The goal is to support restaurant teams—especially back-of-house crews—with shift coverage, prep coordination, and onboarding in a way that genuinely respects workers instead of surveilling them.
This isn’t automation for automation’s sake. It’s not about cutting labor costs or optimizing people into exhaustion. It’s about designing a simple, AI-assisted system that helps small, chaotic teams stay organized—without adding more stress or complexity to already difficult jobs. Having worked in kitchens that used systems like HotSchedules and 7shifts, I’ve seen firsthand how these platforms prioritize management needs while making day-to-day work harder for the people actually on the line.
MEP is meant to do the opposite. It helps assign roles based on real-world context like skill level, fatigue, and task flow—not just raw availability. It can offer onboarding prompts or prep walkthroughs for new cooks during service. Most importantly, it avoids invasive data collection, keeps all AI suggestions overrideable by humans, and pushes for explainability rather than black-box logic.
I’m sharing this here because I want real feedback—not hype. I’m curious how folks in this community think about building AI for environments that are inherently messy, human, and full of unquantifiable nuance. What risks am I not seeing here? What are the ethical or technical red flags I should be more aware of? And do you think AI belongs in this kind of space at all?
This isn’t a startup pitch. I’m not selling anything. I just want to build something my former coworkers would actually want to use—and I want to build it responsibly. Any insights are welcome, especially if you’ve worked on systems in similarly high-stakes, high-pressure fields.
Thanks for your time.
—JohnE
[link] [comments]