Skip to content

Recording a Process

This page walks through a complete recording, from starting the recorder to receiving your finished process.

  • Connect your microphone and make sure it is not muted or in use by another application.
  • The first time, have your ProcessPlan login ready. You use it once to connect the recorder to your account.
  • Close any windows that show confidential or unrelated information. Everything on screen is captured.
  • Plan for a recording of 30 minutes or less. Longer content should be recorded in multiple parts.
  1. Open the recorder window. If it shows Connect to ProcessPlan, click it and approve the connection in your browser, as described in Connecting to ProcessPlan. Once the recorder is connected, the top of the window shows Connected as with your email address. ProcessPlan notifies you at that address when the process is ready.

  2. Choose what the recording becomes:

    • New Process - enter a Process Name. This becomes the process name within ProcessPlan.
    • Merge with Existing Process - search for the process by name and select it from the results. The recording is added to that process.
  3. Select your Microphone. The recorder starts with the microphone Windows is set to use, so if you narrate with a different one, pick it here. Then test it: speak normally and watch the level meter. Your level should consistently reach past the dark vertical line, and the bar should be green rather than red. The line is the same standard the recorder holds you to while recording, so if you clear it here you will not be reminded to speak up later.

    The level meter listens only while the recorder window is the window you are using. Windows shows its microphone icon on the taskbar while it listens, and the icon goes away when you click another window or minimize the recorder. Nothing is recorded until you click Start Recording.

  4. Click Start Recording. The recorder checks its connection to ProcessPlan, the recorder window minimizes to the taskbar, and a notification confirms the recording is active.

The recorder remembers its connection, your microphone choice, and your last process name, so future recordings start faster.

Work at your normal pace in your normal applications. The recorder follows whichever window you are working in, including across multiple monitors.

These keyboard shortcuts work no matter which application you are in:

ShortcutAction
Ctrl+Shift+F10Stop the recording
Ctrl+Shift+F11Pause or resume the recording
Ctrl+Shift+F12Restart the recording
  • Pause when you need to step away or handle something that should not be recorded. Press Ctrl+Shift+F11, hover over the recorder’s taskbar button and click the pause button, or open the recorder window and click Pause. Click Resume (or press the same shortcut) to continue. Paused time does not count toward your recording.
  • Restart if something went wrong and you want a clean take. Press Ctrl+Shift+F12, or open the recorder window and click Restart. The current recording is discarded and a new one begins immediately with the same details.
  • Check your status at any time by hovering over the recorder’s taskbar button, or click it to open the recorder window, which shows the elapsed time and the number of events captured so far. A red dot on the taskbar button means a recording is running. The recorder window is never captured, so opening it does not disturb your recording.
  1. Press Ctrl+Shift+F10, open the recorder window and click Stop, or hover over the recorder’s taskbar button and click the stop button.
  2. The recorder window opens and shows your recording being saved and uploaded, then confirms when the upload succeeds. Once the recording is saved, you can set up your next one while the upload finishes.
  3. ProcessPlan builds the process from your recording. You receive an email when it is ready, at the address shown next to Connected as.

A few things to know:

  • Recordings shorter than 45 seconds are discarded. Very short recordings do not contain enough to build a process from.
  • Recordings where the recorder never heard you at all are discarded. If it could not hear you, it tells you why the recording stopped.
  • If the upload cannot complete (for example, your internet connection dropped), the recording is kept safely on your computer. It is retried automatically the next time the recorder starts, or you can click Retry Upload in the recorder window.
  • If the recorder is interrupted before you stop it - your computer restarts, or the recorder closes unexpectedly - the recording is not lost. The next time the recorder starts it recovers what was captured and uploads it for you.

The process built from your recording follows what you did and said. A few things in it are there on purpose, so you know what they mean when you open it:

  • A step about how the work arrives. Every process starts with the thing that starts the work: an email that comes in, a file that lands in a folder, a report that runs on a schedule. If your recording showed that, the first step describes it. If it did not (for example, the recording began at a login screen), the first step is called How this work arrives and asks you to confirm how the work reaches you.
  • Two linked processes when you handled a list. If you worked through a list one item at a time (tenants on a report, invoices in a mailbox, lines on an order), you get a main process that gathers the list and a second process that handles one item; the main process starts the second one for each item. This is how the process runs many items at once instead of one after another. Anything your recording showed you doing once, after the whole list was finished, such as reconciling a total, sending a summary or filing the batch, stays in the main process and runs once, after every item has been dealt with. The details each item needs are passed from the main process to the per-item one automatically.
  • Steps named “A person handles: …” appear wherever a situation could come up that your recording did not show, such as a record that is not found. They mark where the process needs a person until you say what should happen instead.
  • A list of questions about the process. Open the process, open its settings, and you will find Questions About This Process, with the number still waiting on you in the heading. Each one says what the process assumes and asks you to confirm or correct it: how the work arrives, which items count, what to do when something is missing, how a record is matched when there are several, and which account the process should sign in with. Answers appear underneath the question they belong to, so you and the team can always see what has been settled and what has not. The questions for a single step also appear on that step when you open it in the process designer. See Answering the questions below.
  • Passwords are never kept. If you typed a password or a verification code while recording, it is not stored in the process. The process signs in with its own account, which is set up with you separately.

Answer them by chatting with your AI Strategist - that is the only place your answers are recorded. There are two ways in, and both open the same conversation:

  • From your quote e-mail - click the link to answer the questions about your process.
  • In the app - open the process, open its settings, expand AI Training Videos, and click Answer the questions next to your recording.

If the process was built from more than one recording - for example, you recorded it again and the new recording was added to the same process - your quote and the conversation both cover the open questions from all of its recordings, not just the latest one. The AI Strategist tells you when a question came from a different recording.

The AI Strategist goes through the open questions one at a time, tells you what it assumed, and asks you to confirm or correct it. Each answer is saved against the question it answers, so the team that finishes your process sees what you said, and the question changes from Waiting on you to Answered in the list. If you cannot answer something yet, say so - the question stays open, and your reason is shown beside it so nobody asks you again for the same thing.

A few things to know:

  • Only someone who uploaded one of the process’s recordings, or someone who can edit the process, can answer. Anyone else is told who can.
  • Never type a password, PIN or verification code into the chat. It will not be kept. When the process needs to sign in, you are sent a secure form for that.
  • If an answer changes what was built - for example, the work arrives differently than assumed - the team adjusts the process, and if the price changes you receive a corrected quote.
  • When you say who handles a situation the recording did not show (a record that is not found, a letter that is missing), that step is assigned to the person you name. They receive each case as a task, and once they have handled it the item picks up where it stopped - unless you say the run should stop there instead.
  • When you answer how the work arrives, the process gets its real start: a schedule (it runs itself on the days and at the times you name), a form or link people fill in, an e-mail address your mailbox rule forwards to (one run per message, with the sender, subject, body and attachments in the process’s fields), or a webhook - and the “How this work arrives” step disappears, so every run begins at the first real step. If the work runs at several times of day, say every time it must run: the first one is set up straight away and the rest go to the team with your answer.
  • When you answer how a record is matched (a customer, a vendor, a part), you get a table of mappings - how the value is written on the document and what your system calls it - and a short form. When the process cannot match one record, the person handling it is pointed at that form; the answer is remembered, so the same record is never asked about twice. Any names or aliases you can give right away are saved as the first rows.
  • You can come back to the chat as often as you like. Questions you have already answered are not asked again, and anything still open is waiting whenever you have the answer.

Example documents help. If the process starts from a document or a list, send three to five examples from different senders so the differences between them are covered before the process runs on real work. The easiest way is to attach them in the AI Strategist chat when it asks. If your process gathers a list, the step that reads that list is adjusted from your examples straight away and the Strategist tells you what they showed. If your process handles one item at a time, your examples go to the team, who use them while they finish the process.

If something did not go as expected, see Recorder Troubleshooting.