A CV tells a recruiter what a consultant claims to have done. It does not tell them what to ask to find out if that is true. For a generalist recruiter screening a specialist technical profile, that gap is exactly where a weak interview happens.
AI Prompt Profiles can be configured to close it: translate the CV content that needs to stay exact, and turn each project’s description into a set of technical interview questions the recruiter can actually use. This post walks through how that kind of profile is built and what it produces.
Watch Interview Notes Get Generated Straight from a CV
TL;DR
- An interview-notes profile does two different jobs at once: translate the summary and project description verbatim, and turn each project into a set of technical interview questions.
- The CV summary and project description are never rewritten. They are translated only, sentence by sentence, into the profile’s working language.
- The tasks and responsibilities field gets repurposed. Instead of listing what the consultant did, it holds 8 to 12 technical interview questions for the recruiter to ask about that project.
- The questions are grounded strictly in the project description. Nothing gets asked about a technology or responsibility that is not actually mentioned there.
- Questions are built to separate hands-on experience from surface-level exposure, and yes-or-no questions are explicitly avoided.
- Job title is preserved exactly and returned in English, never invented or reworded.
- Setup happens once in Company Settings. Running it means picking the profile when onboarding a consultant and generating a CV in the interview-notes template.
- The result is a document built for the recruiter, not the candidate, ready to walk into a screening call with.
Why Recruiters Need Interview Questions, Not Just a CV
A consultant’s CV can list Kubernetes, a migration project, and three years on a platform team, and none of that tells a recruiter whether the person configured the cluster themselves or sat next to someone who did. Distinguishing the two usually takes a technical interviewer, and not every recruiting team has one available for every screening call.
The gap shows up most in agencies and staffing firms covering broad technical scopes, where the recruiter running the first screening call is rarely a specialist in the exact stack the project used. They can read the CV. They cannot always tell which question would expose a candidate who is overstating their role.
What an Interview-Notes Prompt Profile Actually Does
Configured for this case, a prompt profile handles the CV summary and project description very differently from the tasks and responsibilities field. The first two are locked to translation. The third is turned into a generation task.
Does this translate the CV, or rewrite it?
Translate only. The summary and project description are moved into the profile’s working language sentence by sentence, with nothing added or reworded.
The instructions behind these two fields are explicit about it: no rewriting, no summarizing, no restructuring, no inferring years of experience, seniority, or achievements. If the source says something plainly, the translation says it plainly, in the target language, and nothing else.
The tasks and responsibilities field works differently on purpose. Instead of listing what the consultant did, it is configured to read the project description and generate a set of technical interview questions about it, using the same underlying AI CV Parser that locates each section on the source CV.
Why the Summary and Project Description Are Translated, Not Rewritten
An interview only works if the recruiter is reacting to what the consultant actually wrote, not to an AI’s interpretation of it. If the summary had been rewritten or embellished along the way, the interview questions generated from it would end up validating a version of the project that never existed.
“Do NOT rewrite, summarise, enhance, restructure, or contextualise the content.”
That instruction appears in both the summary and project description prompts, worded almost identically each time. Translation changes the language. It does not touch the substance, which is exactly what a good interview needs to stay grounded in.
How the Interview Questions Get Generated
The tasks and responsibilities field is where this profile does its most distinctive work. Rather than describing the consultant’s activities, it is instructed to act as an experienced IT recruiter and draft the questions a technical screening interview should ask.
“Focus on questions that help determine whether the candidate actually performed the described work rather than simply being exposed to the technology.”
“Do not ask about technologies, responsibilities, methodologies, or concepts that are not supported by the project description.”
Can the interview questions ask about things that are not on the CV?
No. Every question has to trace back to something explicitly mentioned in the project description.
The prompt is built to avoid generic interview questions that could apply to any candidate. Each question is meant to validate a specific technical claim already present in the project, which is also why yes-or-no questions are avoided in favor of ones that start with “How”, “What”, “Why”, “Which”, or “Can you explain”.
| Field | What it produces | Why it works this way |
|---|---|---|
| CV summary | Verbatim translation into the profile’s working language | Keeps the exact claims the interview will be built around |
| Project description | Verbatim translation into the profile’s working language | Same, at project level |
| Tasks and responsibilities | 8 to 12 technical interview questions grounded in the project | Turns the project description into material the recruiter can actually use |
| Job title / project name | Returned exactly as found, in English | Keeps identifiers consistent regardless of the CV’s original language |
Questions are also written to be understandable to a recruiter who is not a technical specialist, which matters as much as their technical accuracy. A question that only a specialist could parse defeats the purpose of preparing it in advance.
Setting Up and Running an Interview-Notes Profile
Like any AI Prompt Profile, this one is configured once and reused for every consultant onboarded afterward.
- ✓Go to Company Settings and open the AI Review tab
- ✓Click Add AI Prompt Profile and give it a name, for example “Interview Notes”
- ✓Paste the translation prompt into the summary and project description fields
- ✓Paste the question-generation prompt into the tasks and responsibilities field
- ✓Fill in the job title and project name fields to keep those values locked and in English
- ✓Click Save
Running it against a new consultant follows the normal onboarding flow, with one difference in which output template gets selected.
- ✓Click Add a New Consultant and fill in their details as usual
- ✓Check AI Review and select the interview-notes profile from the dropdown
- ✓Check Generate CV and choose the interview-notes template as the output
- ✓Upload the consultant’s existing CV and click Create Account
- ✓Open the generated profile to find the translated summary and project descriptions alongside the interview questions for each project
Companies managing CV management for consulting companies with international teams typically use this to prepare a first screening call the same day a CV comes in, without waiting for a specialist to be free to review it first.
The Actual Interview-Notes Prompts
Here is the real configuration behind each field. Copy any of these into your own AI Prompt Profile as a starting point.
CV Summary Prompt
Instructions:
- Your task is ONLY to translate the content under the "Summary" section of the provided input into {PROFILE_LANGUAGE}.
- Translate the text verbatim, sentence by sentence.
- Do NOT rewrite, summarise, enhance, restructure, or contextualise the content.
- Do NOT calculate, infer, or modify years of experience.
- Do NOT infer roles, achievements, seniority, industries, or career progression.
- Preserve the original meaning, tone, structure, and level of detail.
- Output ONLY the translated Summary text.
- Do NOT output titles, headings, explanations, or any other sections.
- STOP processing after the translation.
Project Description Prompt
Instructions:
- Your task is ONLY to translate the content under the "General Description" section of the provided input into {PROFILE_LANGUAGE}.
- Translate the text verbatim, sentence by sentence.
- Do NOT rewrite, summarise, enhance, restructure, or contextualise the content.
- Do NOT calculate, infer, or modify impact, metrics, seniority, or responsibilities.
- Do NOT introduce or remove technologies.
- Preserve the original meaning, tone, structure, and level of detail.
- Output ONLY the translated General Description content.
- Do NOT output titles, headings, explanations, skills, or any other sections.
- STOP processing after the translation.
Tasks and Responsibilities Prompt
## Objective
Act as an experienced IT recruiter conducting a project discovery and technical validation interview.
Based solely on the information provided in the project description, prepare a list of interview questions designed to:
1. Explore what the candidate actually did in the project, including responsibilities, activities, technologies, decisions, and contributions that may not be explicitly documented.
2. Validate the candidate's real technical experience and understanding of the technologies, architecture, methodologies, responsibilities, and challenges already described.
The interview should not be purely a technical screening. Its purpose is also to discover additional relevant information that could improve the understanding and representation of the candidate's experience.
## Question Balance
Generate approximately 50% exploration questions and 50% technical validation questions.
### Exploration Questions
Exploration questions should uncover relevant experience that may be missing from the project description.
Use the information already provided as a starting point and ask the candidate to explain what they actually did.
Explore areas such as:
* Additional responsibilities they personally performed.
* Specific tasks they handled that are not explicitly listed.
* Technologies or tools they used that are not mentioned in the project description.
* Additional frameworks, platforms, libraries, or services they worked with.
* Other parts of the application or system they contributed to.
* Frontend, backend, cloud, DevOps, architecture, data, testing, security, or other activities they may have performed.
* Their level of hands-on involvement in the project.
* Technical decisions they personally made.
* Problems or challenges they personally solved.
* Integrations, deployments, migrations, troubleshooting, testing, monitoring, or maintenance activities they performed.
* Collaboration with other technical teams or stakeholders.
* Responsibilities they assumed beyond their formal job title.
* Activities they performed that could demonstrate broader expertise relevant to the project.
Exploration questions may go beyond what is explicitly documented, but must remain logically connected to the project and should be designed to discover whether the candidate has additional relevant experience.
Do not assume that the candidate performed any additional activity. Ask questions to discover whether they did.
### Technical Validation Questions
Technical validation questions should assess whether the candidate genuinely understands and performed the technical work already described in the project.
Focus on:
* How technologies were implemented or used.
* Why particular technical approaches were chosen.
* How different technologies interacted.
* Architecture and design decisions.
* Technical challenges and how they were solved.
* Configuration, integration, deployment, testing, troubleshooting, monitoring, or maintenance where relevant.
* Practical trade-offs and alternatives considered.
* The candidate's personal contribution versus work performed by the wider team.
* The candidate's depth of hands-on knowledge.
Prefer practical, experience-based questions over theoretical questions.
The purpose is to distinguish between genuine hands-on experience and superficial exposure to a technology.
## Role and Project Context
Use the project description to understand the candidate's actual context before generating questions.
Consider:
* The candidate's job title.
* The project or system description.
* Explicit responsibilities.
* Technologies and methodologies used.
* The relationship between the technologies.
* The apparent scope of the candidate's involvement.
Prioritise questions that can reveal additional useful information while remaining relevant to this specific project.
Do not turn the interview into a generic technical quiz.
## Question Design
* Each question must have a clear purpose.
* Questions should encourage the candidate to explain their own experience.
* Avoid questions that can be answered simply with "yes" or "no".
* Prefer questions beginning with "How", "What", "Why", "Which", "Can you explain", or "Can you describe".
* Ask about what the candidate personally did rather than what the team generally did.
* Where useful, ask questions such as "What was your specific responsibility?" or "How did you personally approach this?"
* Avoid asking the same question in different ways.
* Do not repeat the same technical topic unless the questions explore substantially different aspects.
* Questions should be understandable to an IT recruiter who may not be a technical specialist.
* Avoid overly theoretical questions unless they are necessary to validate a specific technical claim in the project.
* Avoid generic questions that could apply to any software project.
## Exploring Unlisted Experience
A key purpose of the interview is to discover information that is not currently present in the project description.
When the project suggests a broader area of work, ask questions that can uncover additional technologies, responsibilities, or activities.
For example:
* If backend development is mentioned, explore whether the candidate also worked on APIs, databases, testing, performance, security, monitoring, or deployment.
* If frontend development is mentioned, explore which parts of the UI they implemented, frameworks used, integrations, testing, and user-facing functionality.
* If cloud technologies are mentioned, explore which cloud services they actually used and what they personally configured or managed.
* If architecture is mentioned, explore what they designed, which decisions they made, and the scope of their architectural responsibility.
* If DevOps-related activities are suggested, explore deployment, CI/CD, containers, infrastructure, monitoring, or automation.
* If several technologies are mentioned, explore whether additional tools or technologies were used to connect, support, test, deploy, or operate them.
* If leadership or mentoring is mentioned, explore what they were responsible for and how they contributed to the team.
These questions are intended to discover information, not to assume it.
## Technical Validation
For technical validation questions:
* Focus on technologies and responsibilities explicitly supported by the project.
* Ask how the candidate implemented, configured, integrated, designed, tested, troubleshot, monitored, deployed, or maintained the relevant components where applicable.
* Ask about important technical decisions and the reasoning behind them.
* Ask about challenges encountered and how they were resolved.
* Where multiple technologies are mentioned, explore how they interacted.
* Where an architecture or technical approach is mentioned, validate the candidate's understanding of how it worked in practice.
* Where appropriate, ask about trade-offs, limitations, scalability, reliability, performance, or maintainability.
* Focus on practical experience rather than memorised definitions.
## Factual Boundaries
* The project description is the starting point for the interview.
* Do not present unlisted technologies, responsibilities, or activities as facts.
* Exploration questions may ask about technologies or responsibilities not explicitly listed when they are logically relevant to discovering additional project experience.
* Never assume that the candidate has experience simply because a question asks about it.
* Do not introduce unrelated technologies or generic technical topics.
* Keep all questions relevant to the candidate's project context.
## Output
* Provide 8-12 interview questions.
* Approximately half must be exploration questions and approximately half must be technical validation questions.
* Mix the two types naturally rather than separating them into two sections.
* Use bullet points.
* Each bullet point must contain one question only.
* Do not provide answers, explanations, scoring criteria, or interviewer notes.
* Do not introduce the questions with a header or additional commentary.
* Questions must be written in {PROFILE_LANGUAGE}.
Project Job Title Prompt
If the Job Title is available, return exactly the Job Title. Do NOT summarise or invent a new name. Make sure that is always in English. Return ONLY the job title.
Project Name Prompt
If the Project Name is available, return exactly the Project Name. Do NOT summarise or invent a new name. Make sure that is always in English. Return ONLY the project name.
Why This Matters Beyond One Interview
The same prompt profile is reused for every consultant onboarded with it, in whatever language their original CV was written in. A firm working across several markets can run the same interview-notes profile on a CV written in Portuguese, French, or German and get back translated content plus questions in whichever language the recruiter conducting the call actually works in.
It also sits comfortably alongside a firm’s other AI Prompt Profiles. A consulting company might use a parsing prompt profile for onboarding consultants from a specific framework contract, a job-tailoring prompt profile when submitting for a live opportunity, and an interview-notes prompt profile the moment a CV first lands, before either of the other two ever runs.
Explore More
See What Else an AI Prompt Profile Can Do
Interview notes generation is one configuration. The same capability is also being used to:
See the full picture in AI Prompt Profiles: Match AI CV Review to Every Workflow.
Frequently Asked Questions
What is an interview-notes profile in Sprint CV?
It is an AI Prompt Profile configured to translate a CV’s summary and project descriptions verbatim, while turning each project into a set of technical interview questions for the recruiter.
Does it translate the summary and project description, or rewrite them?
It translates them only. The prompts explicitly forbid rewriting, summarizing, or restructuring the content, and instruct a sentence-by-sentence translation instead.
Are the interview questions for the candidate or the recruiter?
For the recruiter. They are designed to help a screening interviewer verify a candidate’s real, hands-on experience with a project, not to be answered by the candidate in writing.
Can the AI ask about things that are not on the CV?
No. Every question has to be grounded in something explicitly mentioned in the project description. Nothing outside that scope is included.
How many interview questions does it generate per project?
Between 8 and 12, as bullet points, with each bullet containing a single question and no accompanying commentary.
Can I use this alongside a parsing or job-tailoring profile?
Yes. Each is a separate AI Prompt Profile configured for a different case, and a firm can build and use several of them side by side.
Final Thought
Screening a technical consultant should not depend on whether a technical specialist happens to be free for the call. An interview-notes prompt profile puts the groundwork in place ahead of time: the CV content stays exactly as written, translated where needed, and the questions worth asking are already sitting in front of the recruiter.
It is another configuration built on AI Prompt Profiles, alongside parsing and job tailoring, for the specific moment a CV needs to become something a recruiter can act on in the room.
Curious what an AI Prompt Profile built to generate interview notes from your own consultants’ CVs would look like? Book a demo with Marco Pincho to walk through your specific case.
About the Author
Seven years of watching recruiters format the same CV three different ways for three different clients led Marco Pincho to build Sprint CV, an enterprise CV manager used by companies including VASS, Fujitsu, Sword Group, and Expleo to standardize CV creation and submission. He holds an Executive Master in IT Management from Solvay Brussels School and spends most weeks on calls with recruitment leaders and HR directors, working through documentation problems like the one covered in this article. Connect with Marco on LinkedIn
