Docs
Introduction
When creating a project on the platform, you can create a notebook, which is a page where you can write text, code, and mathematical equations. Your entire project is developed and documented within this notebook.
Additionally, next to the notebook, you have access to the HARUMI AI CHAT, our AI assistant that will help you solve your problem.
Step 1 - Documenting the Business Rules
Before even thinking about code or mathematical calculations, the first step on our platform is to provide a detailed description of the context your company wants to optimize — whether it’s a decision-making process, an operation, or logistics, etc.
It’s important to clearly describe the problem to be solved and all its key characteristics. Write this in detail in the notebook so that HARUMI AI can use it as context when formulating the mathematical model.
What’s Important for Building a Model?
You need to clearly define and describe three things in your notebook:
- Variables
- Objective
- Constraints
In mathematics, Variables are quantities whose values you want to determine. In your company’s optimization problem, there may be many variables or unknowns — both quantitative and qualitative — and we aim to find their best values that achieve the Objective while respecting the Constraints.
The Objective refers to what the model is trying to achieve (for example, minimizing costs, maximizing revenue, etc.). It’s expressed as a mathematical equation (involving some or all of the Variables) whose result you want to maximize or minimize.
The Constraints are your business rules translated into mathematical equations for the model. Defining the Constraints is a crucial step in the modeling process. In this case, since our AI helps you build the model and solution, you don’t need to write the equations yourself — just describe the business rule, and the AI will translate it into a mathematical expression for you.
For example, in a vehicle routing problem, a business rule/constraint might be: "A driver cannot drive more than 8 hours per day." In that case, the AI could translate it into the following mathematical constraint:
That is, for each driver k in the problem, the time (in hours) spent driving per day (t_k) must not exceed a maximum allowed time (in this case, 8 hours).
If you want to learn more about defining variables and constraints in an optimization model, read this article from our blog:
What Are Variables and Constraints in Operations Research?
Data
There’s another essential element for executing a mathematical model properly: the data of your problem. Not every term in the equations (either in the Objective Function or in the Constraints) is a Variable to be determined. Some values must be provided by you through spreadsheets or similar files.
Details on how to upload data and configure them to be recognized within the notebook will be explained in the next steps.
Step 2 - Uploading Data to the Platform
What you need to do is add your file to the project using the Files section on the left side. Once added, you can set the file as a parameter in your notebook.
See an example at:
[TUTORIAL] Uploading Production Planning and Control (PPC) Data
Step 3 - Parameter Configuration
- With your notebook opened, click on Parameters to open the parameters section
- Click on + New Parameter, and then click on File.
- A new row will appear in the parameter table. Simply select the file you previously added and give it a name. With this, you’ll have a “variable” globally defined within your code, and your notebook will recognize the file by the name you assigned in the Parameters table.
