Integration

Integrating with Timefold Platform gives you access to Timefold’s solver technology which allows you to:

  • Solve your planning problems and optimize your plans.

  • Respond to change with recommendations and real-time planning.

  • Receive insights about your planning problems.

  • Analyze and compare your solutions.

1. API

Timefold Platform is integrated with your existing systems through our API.

  1. Your systems create input datasets with your problem data that is sent to Timefold by the API.

  2. Timefold solves the input dataset and sends back an output dataset with the optimized results, input metrics, and output metrics.

This approach means you maintain control over your data. The planning data and planning module remain part of your own backend system, or they could be embedded as a CRM module.

2. Basic workflow

  1. Your system gathers the problem data.

  2. Your system make a POST call sending the planning data (input dataset) to Timefold.

  3. Timefold optimizes the plan.

  4. Timefold sends the solution (output dataset).

  5. Your system processes the results.

You can see all of the datasets that have been queued for solving in the solve queue.

3. Waiting for completion

Solving is an asynchronous process which might take several minutes and in some cases even hours depending on your configuration. In order to check the status of an optimization you have 3 options, each with their own trade-offs.

Mechanism Key Advantages Key trade-offs

Webhooks

Real-time, efficient, auditable (via Platform UI)

Public endpoint, security & retries

Server Sent Events (SSE)

Simple real-time push

One-way, long-lived connections

Polling

Simple, universal

Inefficient

4. Responding to change

Plans are rarely executed without modifications as it is often necessary to respond to real-world events, such as a customer being suddenly unavailable or an employee calling in sick.

When plans need to be updated, you can update your original input dataset (or the latest version if this is not the first change) with the new data.

  1. Your system registers there has been a change.

  2. Your system make a POST call sending updated planning data (input dataset) to Timefold.

  3. Timefold optimizes the plan.

  4. Timefold sends the solution (output dataset).

  5. Your system processes the results.

5. Working with recommendations

When you need help figuring out which employee can cover a shift at the last minute or which technician to send to an emergency visit, you can use the recommendation engine.

Recommendations don’t re-optimize an existing plan, but they do help your support staff offer customers options, which once accepted can optionally be added to an updated plan:

  1. Your system registers there has been a request for a change.

  2. Your system makes a POST call to the recommendation endpoint with the ID of the entity it needs a recommendation for.

  3. Timefold send back a list of recommendations.

  4. (Optional) Your system make a POST call sending updated planning data (input dataset) to Timefold.

  5. Timefold optimizes the plan.

  6. Timefold sends the solution (output dataset).

  7. Your system processes the results.