Supply Chain and Logistics Business Ideas

"Define and drive technical strategy, architecture standards, and engineering roadmaps for next-generation AI-powered systems Lead the design and architecture of enterprise-scale AI, agentic, and automation platforms that drive measurable business outcomes."

"Some part of it is billing API maturity, believe or not. I worked on model distribution for a big tech and we didn't have a robust billing API in 2024/25. Models were coming out faster than reinforcing the infra layer and securing GPU was the primary focus. We fought for it but couldn't get it funded. So providing the most transparent billing was...not possible in some regard. We refunded quite a number of cases where customer accidentally occurred a huge bill because they didn't know what they were doing. As a PM, it was a tough bill to swallow. Also, historically, top-up billing was not considered in the Enterprise SaaS as a dark pattern because renewal friction was really hard in the classic ML days where the team lead had to negotiate with the IT/finances. Some ML teams gamed it by spending upfront before others can use it as compute was a limited resource. Giving a team a buffer while finances can purchase additional SKU/token was seen as an user experience improvement. In some cases, spending limit was a *requested* feature by IT teams. Consumer use case is of course different but I was always so curious how quickly startups were getting the billing up and running. Accurate token measurement in a dynamic pricing (PTU, bundle pricing, etc) is not an easy feat. So either it was a very simple billing model or someone was manually running the billing numbers somewhere for high paying customers while first party/third party processors were catching up."

"Like are you holding on to inventory that you should be liquidating faster, right? Because you can still, you know, and I think there's especially in retail, a lot of people are like, well, you don't want to have markdowns. I'm like, but if you, like, you can still make money depending on your margins. If you mark it down 20% and it's going to move it faster, you still make money at 20% off, right? You make slightly less money, but the implications of holding too much of that bad inventory hurt you further on down the road. So you have to think about what's performing, but also what's not performing and clearing your, like, discontinuing it and getting rid of the stuff that's not driving your business because that's just going to weigh you down."