Asking Questions About Your Interview Data

Overview

candidate.fyi lets you ask questions about your interview and scheduling data using natural language. Instead of manually building filters or exports, you can simply type a question like "How many interviews were conducted last month?" and get an answer.

This feature is available in the Insights section of candidate.fyi and helps you quickly access scheduling metrics, interviewer activity, and candidate engagement data without learning complex reporting tools.

What are semantic queries?

Semantic queries are natural language questions you can ask about your recruiting data. The system understands what you're asking and retrieves the relevant information from your interview and scheduling records.

For example, instead of filtering by date ranges, selecting columns, and exporting a report, you can ask:

• "Get the total interviews conducted this month"

• "Interview count this year broken down by scheduler"

• "Reschedule rate broken down by department"

The system interprets your question, finds the data, and presents it in a clear format.

Why use semantic queries?

Semantic queries help you:

• Save time — Get answers in seconds without building custom reports.

• Access data naturally — Ask questions the way you think about them, without learning query syntax or navigating complex menus.

• Make faster decisions — Pull metrics on demand during planning meetings or one-on-ones.

• Spot trends quickly — Compare performance across schedulers, departments, jobs, or time periods with a simple question.

Who can use semantic queries?

This feature is available to organization admins and editors. If you do not see the semantic query interface, check your role permissions in candidate.fyi.

Understanding Ask mode and Build mode

The semantic query interface has two modes with the same underlying capabilities. The difference is how they handle your queries:

Ask mode

Ask mode is for exploring your data and learning about what you can do. When you ask a question:

• The system answers your question and shows you the data.

• It does not create or modify any saved metrics.

• It does not change any visualizations you're working on.

• Use this mode to get quick answers or learn about the Ask fyi agent's capabilities.

Build mode

Build mode is for creating and managing metrics. When you ask a question or make a request:

• The system answers your question and shows you the data.

• It creates or updates the metric you're working on.

• It changes the visualization based on your request.

• Use this mode when you want to save a metric or build a visualization.

Both modes have the same capabilities—they understand the same questions and can produce the same results. The only difference is whether your work gets saved as a metric or remains a one-time answer.

Switching between modes

You can switch between modes at any time:

• Use the mode toggle in the interface, or

• Press Cmd+Shift+M (Mac) or Ctrl+Shift+M (Windows/Linux)

How to use semantic queries

1. Navigate to the Ask fyi page in the Insights section of candidate.fyi.

2. Choose Ask mode for quick questions or Build mode if you want to create a metric.

3. Type your question in plain language in the chat interface.

4. Review the results. You can toggle between summary views and raw data.

If the results don't match your expectations, refine your question and try again.

Example questions you can ask

Here are some common semantic queries to help you get started:

Interview volume

• "Get the total interviews conducted this month"

• "Interview count this year broken down by scheduler"

• "Get the interview count for last quarter broken down by month"

• "Get the total interviews scheduled for [scheduler's name]"

Scheduling efficiency

• "Get the time to schedule by scheduler"

• "Get the average time to schedule by job"

• "Reschedule rate broken down by department"

Custom breakdowns

You can ask for breakdowns by:

• Scheduler

• Department

• Job or role

• Interview round

• Time period (week, month, quarter, year)

Saving and sharing queries

You can save your queries as metrics for future reference when you're on the create metric or edit metric page:

• Save your query — Saved metrics appear in your personal list and can be updated as needed. Only you can edit your own saved queries.

• Visibility — By default, saved metrics are only visible to you. You can choose the Publish option to make them public so other team members can view and use them.

• Copy queries — You can copy your own saved queries or queries created by other team members to reuse or modify them.

• View existing metrics — Navigate to the Metrics page to browse all saved metrics, including your own and those shared by others.

This makes it easy to build a library of frequently used reports and share insights across your team.

Tips for writing effective queries

Reference people, jobs, and departments directly

When asking about specific people, jobs, or departments, you can reference them directly without any special symbols like @ or [ ]:

• Schedulers and interviewers — Use their name or email address. For example: "Get interview count for Jane Smith" or "Show me availability for jane.smith@company.com"

• Jobs — Use the job name or job code. For example: "Interviews for Senior Engineer" or "Time to schedule for job code ENG-123"

• Departments — Use the department name. For example: "Reschedule rate for Engineering"

Simply type the name or identifier as you would say it naturally—no special formatting required.

Use consistent terminology

Refer to people, jobs, and departments the way they appear in candidate.fyi. For example, if a scheduler is listed as "Jane Smith," use that exact name in your query.

Ask for one thing at a time

Instead of "How many interviews were scheduled last month and what was the reschedule rate and who were the most active interviewers," break it into three separate questions:

• "How many interviews were scheduled last month?"

• "What was the reschedule rate last month?"

• "Who were the most active interviewers last month?"

Request breakdowns when helpful

If you want to compare performance across groups, include "broken down by" in your query:

• "Interview count broken down by department"

• "Time to schedule broken down by scheduler"

Start simple

If you're new to semantic queries, start with straightforward questions like "How many interviews were conducted last month?" Once you're comfortable, you can ask more detailed questions.

Understanding your results

Results may include:

• Summary metrics — A single number or percentage (for example, "Total interviews: 245").

• Breakdowns — Tables showing metrics grouped by scheduler, department, or job.

• Time-series data — Trends over weeks, months, or quarters.

• Raw data — Detailed records you can export for further analysis.

You can toggle between views depending on the question and your needs.

Visualizing your data

You can ask for your results to be displayed in multiple formats:

• Table — The default view showing data in rows and columns.

• Line chart — Useful for showing trends over time.

• Bar chart — Good for comparing values across categories.

• Pie chart — Shows proportions and percentages of a whole.

• Multi-line chart — Compares multiple trends over time on the same chart.

• Grouped bar chart — Compares multiple categories side by side.

Simply include your preferred visualization in your query, such as "Show me interview count by month as a line chart" or ask the system to visualize the results after receiving them.

Troubleshooting

My results seem incorrect

• Double-check the time range, names, or terminology in your query.

• Make sure the data is synced from your ATS and calendar integrations.

• Contact candidate.fyi support with details about what you expected versus what you received.

The system doesn't understand my question

• Simplify your question and remove extra words.

• Use terminology that matches how data appears in candidate.fyi (for example, scheduler names, job titles, department names).

• Try rephrasing using examples from this article.

Getting help

If you have trouble with semantic queries or need help crafting a specific question, contact candidate.fyi support. Include:

• The question you asked

• What you expected to see

• What results you received (or if you received an error)

Our team can help refine your query or identify any data issues.