1. Custom Questions (Manual)
Custom Questions are static, pre-written questions that your team explicitly authors. The AI asks these exactly as programmed for every candidate.When to use them
- Behavioral interviews: “Tell me about a time you handled a difficult client.”
- Core competencies: To ensure absolute consistency, every candidate receives the exact same fundamental question.
- Predictable scoring: You clearly define what constitutes a “Strong (5)” and “Weak (1)” answer in the rubric.
Key characteristics
- Static input: You write the question manually.
- Strict rubric: Requires specific benchmarks (
looking_for,strong_answer,weak_answer,weight). - Predictable consistency: The AI does not improvise the base question, keeping the interview track completely uniform.
2. Resume Questions (Dynamic)
Resume Questions rely on the AI’s ability to instantly read and analyze the candidate’s uploaded resume. Instead of asking a generic question, the AI generates questions tailored precisely to the candidate’s past work history.When to use them
- Verifying experience: Deep-diving into specific projects or promotions listed on standard resumes.
- Tailored exploration: Discovering the depth of a candidate’s actual responsibilities versus what they simply listed on paper.
Key characteristics
- Dynamic input: The AI generates the question on the fly based on the resume text and your
generationInstructions(up to 4,000 characters). - Highly personalized: No two candidates will receive the exact same question.
- Follow-up intensive: You can instruct the AI to heavily scrutinize gaps in employment or specific project claims.
3. Skills Questions (Dynamic)
Skills form the foundation for evaluating competencies. When configuring an interview, you can add skills—such as Communication, System Design, or Problem Solving—to focus the AI’s questioning. The interview can be as complex or as straightforward as you need it to be. Because a single skill like “Communication” is incredibly broad and varies drastically depending on the role, adjusting this shared instruction is key to receiving the exact insights you want. See below how refining the common instruction for a “Problem Solving” skill transforms the resulting interview experience:Step 1: Broad Assessment
The most basic and open type of evaluation.- Instruction:
(no instructions) - Example Questions generated by AI:
- “Describe a time when you had to solve a difficult problem at work without having all the information upfront.”
- “How do you approach a situation where your initial solution to a problem completely fails?”
- “Tell me about a time you had to analyze a complex issue and come up with an unconventional solution.”
Step 2: Adding Context
Now, let’s specify what kind of problem solving matters for this role.- Instruction:
Evaluates the candidate's ability to resolve angry customer complaints for a software product. - Example Questions generated by AI:
- “Describe a time you had to successfully de-escalate a situation with an incredibly frustrated user who experienced a software bug.”
- “How do you handle a customer who demands a feature that your platform fundamentally does not support?”
- “Tell me about a time you had to turn a negative customer experience into a positive long-term relationship.”
Step 3: Modifying Tone and Length
Let’s ask InterviewFlowAI to ask shorter, more direct questions.- Instruction:
Evaluates the candidate's ability to resolve angry customer complaints for a software product. Keep questions under two sentences. - Example Questions generated by AI:
- “How do you quickly calm down an enterprise user who is furious about a major service outage?”
- “Tell me exactly how you would handle a client requesting a full refund due to a known system bug.”
- “What is your step-by-step process for de-escalating a frustrated customer on a live call?”
Step 4: Providing Scenarios
Providing context through examples helps candidates understand exactly what you are looking for.- Instruction:
Evaluates the candidate's ability to resolve angry customer complaints for a software product. Keep questions under two sentences and use a role-play scenario. - Example Questions generated by AI:
- “Imagine I am an enterprise client threatening to cancel our contract because of a recent system crash. How would you handle my complaint right now?”
- “A user has been waiting for a critical bug fix for two months. How do you respond to their angry email demanding an immediate timeline?”
- “If a customer is screaming on the phone because they lost data due to a user error, what is the very first thing you say to them?”
Summary Comparison
| Feature | Custom Questions | Resume Questions | Skills Questions |
|---|---|---|---|
| Origin | Manually written by you | Dynamically generated from the resume | Dynamically generated from provided skills and a common instruction |
| Consistency | Exact same phrasing per candidate | Highly customized per candidate | Tailors dynamic dialogue to match the desired competency level |
| Best used for | Core competencies, behavioral | Drilling into past experience | Technical, strategic, or behavioral skill evaluation |