More and more of our daily lives rely on logistics automation.
Just think about it, you may have ordered Grubhub for dinner last night, Instacart for your groceries, or Takl for your home repair. In all of these cases, your experience is determined by an algorithm (often many!).
These may seem like simple systems. Just assign the closest driver to the next delivery, right? Wrong. These decision models take in tons of information to make sure not only you have a good experience but all the other users do too. If you listen to tech news, you know that margins in logistics are really low and automation is core to these on demand platforms as they scale.
That’s why we created Hop, to help you build algorithms the same way you build other software - with simple, interpretable computer code.
Hop is a decision modeling and optimization tool built for developers. It helps you automate decisions like routing, scheduling, and assignment.
Hop helps you put automated decisions in place quickly, iterate on them easily as your requirements change, and automatically gather evidence along the way so you know if it's working the way you want. This minimizes time spent on infrastructure and maximizes your impact as a modeler.
Hop is unique in that it encourages users to follow decision engineering best practices. An automated decision should be:
- Repeatable
- Testable
- Interpretable
Repeatable decisions are easy to create and deploy to different environments. They are configured the same way in research and development, testing, and deployment. They can re-use business logic so you don't violate the DRY principle (don't repeat yourself).
Testable models are easy to validate and configure in CI/CD. Recoding decision logic into an unfamiliar language (e.g. linear inequality systems) introduces layers of potential errors. Hop models are built from state. This makes testing decisions like testing other software.
Interpretable decisions don't require complex data transformation once a decision is made. Decisions can be stored and queried directly just like any other data. They can be picked up and re-created from their inputs for debugging. The modeler owns the model. It's not a black box behind a REST API.
We help you get automated systems off the ground fast and support with experimentation tools like simulation to demonstrate performance. This lets you speak the same language as your stakeholders - KPIs and profitability.
We launched Hop in December and our first user developed an optimization algorithm for minimizing delivery time in less than 24 hours.
We’re nextmv and we help you build algorithms in days vs months.
