# DIGIT-TM II to DIGIT-TM III CV Conversion with AI Prompt Profiles

> See how to configure an AI Prompt Profile in Sprint CV to convert DIGIT-TM II CVs into DIGIT-TM III format.

- URL: https://www.sprintcv.com/blog/digit-tm-ii-to-digit-tm-iii-cv-conversion-with-ai-prompt-profiles/
- Language: en
- Author: Marco Pincho, Founder and CEO
- Published: 2026-09-03
- Last updated: 2026-09-03
- Publisher: Sprint CV (https://www.sprintcv.com/)

---

Companies working under the European Commission’s DIGIT-TM framework contract are in the middle of a template change. DIGIT-TM III replaces DIGIT-TM II, and every consultant CV that reaches your team in the old format has to be rebuilt in the new one, section by section, without losing the wording a client or contracting authority has already accepted.

[AI Prompt Profiles](https://www.sprintcv.com/blog/ai-prompt-profiles-match-ai-cv-review-to-every-workflow/) is the Sprint CV capability built for exactly this kind of case: you configure a profile with instructions for a specific CV format, then reuse it for every consultant who arrives in that format. This post walks through how to configure one for the DIGIT-TM II to DIGIT-TM III conversion specifically, what it should extract, where it should draw the line between one section and the next, and what it should and should not touch along the way.

See it in action

Watch the DIGIT-TM Parsing Profile Get Built, Then Set Up Your Own

[Watch on YouTube](https://youtu.be/iY0jRihgOpk) [Read the Step-by-Step](https://www.sprintcv.com/blog/how-to-create-and-use-an-ai-prompt-profile/)

**TL;DR**

- AI Prompt Profiles can be configured to handle the DIGIT-TM II to DIGIT-TM III migration: one named profile, built once, reused for every consultant onboarded under it.
- Configured this way, it locates three sections on a DIGIT-TM II CV, the summary, the project description, and the roles and responsibilities, and copies each one into the matching DIGIT-TM III field exactly as written.
- Nothing is rewritten, reworded, reordered, or corrected. The profile stops the instant it reaches the next section heading.
- It can be set up to recognize both CV variants DIGIT-TM allows, whether the roles and responsibilities section is written for an employee or for an external service provider.
- Template labels sitting at the top of a project description are removed automatically, without touching the actual content behind them.
- Two fields can work differently from the rest: job title and project name capped at three words and only translated if the target language requires it.
- If an expected section is missing, a well-configured profile returns nothing for that field rather than guessing or inventing content.
- The same logic extends to any other framework contract with a fixed CV template, each as its own profile.

In this article

1. Why DIGIT-TM II to DIGIT-TM III conversion is a real onboarding problem
2. What a DIGIT-TM II to DIGIT-TM III profile should extract
3. How to decide where a section starts and ends
4. What gets cleaned up, and what never changes
5. What happens when a section is missing
6. Setting up a DIGIT-TM II to DIGIT-TM III profile
7. The Actual DIGIT-TM II to DIGIT-TM III Prompts
8. Why this matters beyond DIGIT-TM
9. Frequently Asked Questions

## Why DIGIT-TM II to DIGIT-TM III Conversion Is a Real Onboarding Problem

[DIGIT-TM is the European Commission’s framework contract](https://www.sprintcv.com/blog/digit-tm-iii-framework-contract-the-new-european-commissions-it-services-procurement/) for IT consultancy, and firms delivering under it do not get to choose their CV template. When the contract moves from DIGIT-TM II to DIGIT-TM III, every consultant CV on file, and every new one that arrives during onboarding, has to be restructured to match.

Done by hand, that is dozens or hundreds of CVs rebuilt one at a time, by people who also have deadlines, submissions, and interviews to manage. Done carelessly, with a generic AI cleanup step, it creates a different problem entirely.

That is the actual risk with treating a DIGIT-TM II to DIGIT-TM III conversion like any other CV import. The content has typically already been reviewed and accepted once. Rewriting it, even lightly, invites a conversation nobody on the delivery team has time for.

The pressure is rarely about a single CV. It shows up when a firm wins a large lot under the contract and has to onboard a batch of consultants at once, or when an existing team needs every CV on file updated before a submission deadline. A conversion that takes a recruiter fifteen minutes per CV becomes a multi-day project the moment volume enters the picture.

## What a DIGIT-TM II to DIGIT-TM III Profile Should Extract

Configured for this case, a profile is set up to pull three specific blocks of content from a source CV and place each one, unedited, into its matching DIGIT-TM III field.

What exactly should this kind of profile copy?

Three sections: the CV summary, the project description, and the roles and responsibilities for each project.

The summary is located and copied in full, sentence for sentence. The project description is taken from the section DIGIT-TM II labels as the general description of each assignment. The roles and responsibilities text is taken from the section describing what the consultant actually did on the project. All three are placed into the corresponding DIGIT-TM III fields with the wording, punctuation, and line breaks intact.

Underneath, this runs on the same [AI CV Parser](https://www.sprintcv.com/ai-cv-parser/) that powers every AI Prompt Profile, configured here specifically for the layout DIGIT-TM II documents follow.

## How to Decide Where a Section Starts and Ends

A DIGIT-TM II document is structured, but structure alone does not tell an AI where one section stops and the next begins. The profile needs explicit boundary rules for each field, so it never pulls in a paragraph from the wrong section.

| DIGIT-TM II source section | Maps to | Stops before |
| --- | --- | --- |
| Summary | CV summary | The next top-level section (Experience, Education, Skills, and so on) |
| General Description | Project description | The roles and responsibilities heading |
| Employee’s, or external service provider’s, roles & responsibilities | Tasks and responsibilities | The technologies and methodologies heading |

DIGIT-TM II CVs are written for two kinds of contributors, an employee or an external service provider, and the section headings differ slightly depending on which one authored the document. Worth building the profile to recognize both versions of the roles and responsibilities heading, and both versions of the technologies heading that follows it, so it works regardless of which variant a given CV uses.

That distinction matters more than it might look. Staffing and consulting firms delivering under DIGIT-TM regularly mix internal employees with subcontracted consultants on the same lot, and the two groups’ CVs are not always formatted identically going in. A profile that only recognized one heading variant would quietly fail on every CV from the other group, and someone would have to notice the gap before it became a problem in a submission.

## What Gets Cleaned Up, and What Never Changes

DIGIT-TM II project descriptions often carry repeated boilerplate at the top, a “Project description” label, sometimes a “Project size” line above it, before the actual description begins. It is worth telling the profile to strip that boilerplate only when it appears at the very start of the extracted text. Anything that looks similar further down in the body should be left exactly as written, since removing it there would mean guessing at intent rather than following a fixed rule.

Should a profile like this ever translate or edit anything?

At most two fields: job title and project name. Everything else should be copied exactly as written.

Job title and project name can be capped at three words and translated into the CV’s working language only when that language actually requires it, with the original wording preferred whenever it can be kept. The summary, project description, and roles and responsibilities should never go through this step. They are extracted, not composed.

## What Happens When a Section Is Missing

Not every CV that reaches the system will have every section filled in, and a well-configured profile handles that without improvising. If an expected heading cannot be found, or there is nothing between two headings that should contain content, the field should be left empty instead of being filled with a guess.

- ✓Extraction only happens between the exact headings the profile is configured to look for
- ✓Nothing is invented when a section is missing or empty
- ✓Boilerplate is only removed from the very start of the extracted text, never from further inside it
- ✓Both employee and external service provider heading variants are recognized as equivalent
- ✓Job title and project name are the only fields that can be shortened or translated
- ✓The summary, project description, and roles and responsibilities are never rewritten, reordered, or corrected

For a bid or tender manager, that distinction matters as much as the extraction itself. A field that is empty is easy to spot and fix. A field quietly filled with invented text is not, and in a compliance-driven contract like DIGIT-TM, that is the kind of error nobody wants to catch after submission.

What if a CV does not follow the DIGIT-TM II template exactly?

Whichever fields depend on the missing or reworded heading come back empty. The rest of the CV still loads normally.

A CV that has been edited outside the standard template, with a renamed heading or a merged section, will not break the import. It simply means the profile cannot locate that specific block of text, so the field is left for the recruiter to fill in manually rather than being populated with a mismatch. Everything else on the CV that does follow the template is still extracted as expected.

## Setting Up a DIGIT-TM II to DIGIT-TM III Profile

Like any AI Prompt Profile, this one is configured once, in Company Settings, and then reused for every consultant who arrives with a DIGIT-TM II CV.

- ✓Go to Company Settings and open AI Review
- ✓Click Add AI Prompt Profile and give it a clear name, for example “DIGIT-TM II Parser”
- ✓Fill in the prompt fields
- ✓Save the profile
- ✓Click Add a New Consultant and fill in their details as usual
- ✓Check AI Review, then pick the profile you just built from the AI Prompt Profile dropdown
- ✓Check Generate CV, choose DIGIT-TM III as the output template, and pick a file format
- ✓Upload the consultant’s existing DIGIT-TM II CV and click Create Account

Once saved, the AI reads each uploaded file using that profile and builds the consultant’s record with summary, project description, and responsibilities placed exactly as they were written, while producing a ready DIGIT-TM III CV in the same step. Firms managing [CV management for consulting companies](https://www.sprintcv.com/enterprise-cv-manager/) at volume typically run this once per batch of consultants onboarded under the same contract.

## The Actual DIGIT-TM II to DIGIT-TM III Prompts

Rather than describe the prompt fields in the abstract, here is the real configuration, field by field. Copy any of these directly into your own AI Prompt Profile as a starting point, and adjust the headings if your source CVs use slightly different wording.

**CV Summary Prompt**

```
You are an information extraction engine.

Your task is to extract the Summary section from the input CV exactly as it appears.

Instructions
Locate the section titled Summary.
Return only the content belonging to that section.
Stop when the next top-level section begins (for example: Experience, Education, Skills, Certifications, Languages, Projects, etc.).
Preserve:
Every sentence
Every paragraph
Original wording
Original punctuation
Original capitalization
Original line breaks
Do NOT:
Rewrite
Improve
Correct grammar
Correct spelling
Translate
Reorder sentences
Remove repeated text
Add introductory text
Add concluding text
Add Markdown formatting
If no Summary section exists, return an empty string.

Output
Return only the extracted Summary text.
```

**Project Description Prompt**

```
You are extracting information from a DIGIT-TM2 CV.

The document always follows the DIGIT-TM2 template.

Your task is to extract only the content of the General Description section.

Extraction Rules

Locate the exact heading:
General Description

Extract every line that appears after this heading.

Stop extracting immediately before the exact heading:
Employee's Roles & Responsibilities in the project:

Normalization Rules

After extracting the content, perform the following normalization steps only at the beginning of the extracted text.

Step 1
Inspect the first extracted line.
If it is exactly one of the following:
*Project description*
*Project description:*
Project description
Project description:
Project description :
remove that first line.

Step 2
After Step 1 (if applicable), repeatedly inspect the beginning of the extracted content.
If the first line starts with:
Project size:
remove that line and also remove every subsequent line until a line is found that is exactly one of:
Project description
Project description:
Project description :
*Project description*
*Project description:*
Then remove that Project description line as well.

Repeat this process if necessary until the extracted text no longer begins with either a Project size section or a Project description heading.

Only perform these removals when they occur at the very beginning of the extracted content.
Do not remove any occurrence appearing later in the text.
Do not remove any other text.

Preservation Rules

The extracted text must be identical to the original document, except for the normalization rules described above.

Preserve exactly:
- bullet points
- project numbering (e.g. Project 1:)
- indentation
- paragraphs
- punctuation
- capitalization
- wording
- line breaks

Do Not
- rewrite
- summarize
- improve
- interpret
- infer
- translate
- merge or split paragraphs
- merge or split bullet points
- remove duplicated information
- correct grammar or spelling
- add Markdown formatting
- add explanations
- add introductory or concluding text

Validation

If the heading General Description is not found, return an empty string.
If the heading Employee's Roles & Responsibilities in the project: is not found, return an empty string.
If no content exists between the two headings, return an empty string.

Output
Return only the extracted General Description, preserving the original formatting except for the normalization rules above.
```

**Tasks and Responsibilities Prompt**

```
You are an information extraction engine parsing a DIGIT-TM2 Curriculum Vitae.

Your task is to extract ONLY the content of the project's roles and responsibilities section.

SECTION TO EXTRACT

The target section may use either of the following headings:
1. Employee's Roles & Responsibilities in the project:
2. External service provider's roles & responsibilities in the project:

Treat both headings as equivalent. Extract the content belonging to whichever heading is present.

SECTION BOUNDARIES

The target section begins immediately after the target section heading.

The target section ends immediately before the following section heading:
Technologies and methodologies used by the employee in the project:

The following heading may also appear instead:
Technologies and methodologies used by the external service provider in the project:

Treat both technology-section headings as equivalent section boundaries.

IMPORTANT EXTRACTION RULES

- Never include the target section heading in the output.
- Never include either version of the target section heading in the output.
- Never include the following Technologies and methodologies section or its heading in the output.
- Extract only the text between the target section heading and the following Technologies and methodologies section heading.
- Do not extract text from the Project name, Job title, Company, Dates, Client, Project size, Project description, General Description, Technologies, Skills, or any other section.
- Preserve the extracted content exactly as it appears in the source document.

PRESERVATION RULES

Preserve exactly:
- Original wording
- Original spelling
- Original capitalization
- Original punctuation
- Original paragraph structure
- Original bullet points
- Original line breaks
- Original numbering
- Original indentation
- Original Markdown formatting, if present
- Original labels such as "Role:"
- Original whitespace within lines where it is part of the extracted text

Do NOT:
- Rewrite
- Summarize
- Improve
- Correct grammar
- Correct spelling
- Translate
- Rephrase
- Interpret
- Infer
- Add information
- Remove information
- Reorder content
- Merge paragraphs
- Split paragraphs
- Merge bullet points
- Split bullet points
- Remove duplicated text
- Add headings
- Add labels
- Add explanations
- Add introductory or concluding text
- Add Markdown formatting that was not present in the source

OUTPUT RULES

- Return ONLY the extracted body of the target section.
- Do not include quotation marks around the output.
- Do not include the section heading.
- Do not include any explanation or commentary.
- Do not include the Technologies and methodologies section.
- Do not include any text before or after the extracted content.

VALIDATION

Before returning the answer, verify:
1. The output starts with the first content belonging to the target roles and responsibilities section.
2. The output ends immediately before the Technologies and methodologies section.
3. Neither section heading is included.
4. No content from another section has been included.
5. The original wording and formatting have been preserved.

If the target roles and responsibilities heading is not found, return an empty string.
If the corresponding Technologies and methodologies heading is not found, return an empty string.
If there is no content between the two section headings, return an empty string.

Return only the final extracted text.
```

**Project Job Title Prompt**

```
Write the job title in {PROFILE_LANGUAGE}, using no more than 3 words.

Prefer to preserve the original job title exactly as written whenever possible. Only translate it if required by {PROFILE_LANGUAGE}. If no job title exists, generate a concise and accurate job title.

Do not include any header, label, explanation, or quotation marks. Output only the job title.
```

**Project Name Prompt**

```
Write the project name in {PROFILE_LANGUAGE}, using no more than 3 words.

If the original project name exists, preserve it as closely as possible. If no project name exists, generate a concise descriptive name.

Do not include any header, label, explanation, or quotation marks. Output only the project name.
```

## Why This Matters Beyond DIGIT-TM

This DIGIT-TM II to DIGIT-TM III setup is one configuration of a broader capability inside Sprint CV, AI Prompt Profiles, which lets a team define exactly how the AI should handle any given CV format. The same section-boundary logic used here, locate a heading, extract until the next one, leave everything else untouched, can be rebuilt as its own profile for ITS onboarding, another framework contract, or a client’s own internal CV template.

Building a new profile from scratch follows the exact structure described above: pick the source headings, decide where each section stops, and note which fields, if any, are allowed to be shortened or translated. A firm that has already mapped its DIGIT-TM II to DIGIT-TM III conversion has effectively written the template for its next one.

Firms evaluating whether this level of control is worth setting up can see the broader case in [Enterprise CV Manager](https://www.sprintcv.com/enterprise-cv-manager/), or look at how other consulting and staffing teams use it in [Enterprise Testimonials](https://www.sprintcv.com/enterprise-cv-manager/).

Explore More

See What Else an AI Prompt Profile Can Do

DIGIT-TM conversion is one configuration. The same capability is also being used to:

[→ Tailor a CV to a job description Reorders a consultant’s real experience around a job description, without inventing new skills.](https://www.sprintcv.com/blog/tailor-a-cv-to-a-job-description-with-ai-prompts/) [→ Generate interview notes from a CV Translates summaries verbatim and drafts technical screening questions per project.](https://www.sprintcv.com/blog/generate-interview-notes-from-a-cv-using-ai-prompts/)

See the full picture in [AI Prompt Profiles: Match AI CV Review to Every Workflow](https://www.sprintcv.com/blog/ai-prompt-profiles-match-ai-cv-review-to-every-workflow/).

## Frequently Asked Questions

**Is there a dedicated DIGIT-TM II Parser tool in Sprint CV?**

Not as a separate product. It is an AI Prompt Profile, the same general capability used for every other case, configured with the specific headings and rules DIGIT-TM II to DIGIT-TM III conversion needs. You build and name it yourself, for example calling it “DIGIT-TM II Parser.”

**Does it change any wording on my consultants’ CVs?**

No, when configured as described here. The three core fields extracted, summary, project description, and roles and responsibilities, are copied exactly as written. Only job title and project name can be shortened or translated, and only when necessary.

**What happens if a CV is missing a Summary or General Description section?**

A well-configured profile returns an empty field rather than generating placeholder text. It only extracts content that actually exists between the headings it is looking for.

**Can it handle CVs written for external service providers as well as employees?**

Yes, if the profile is built to. DIGIT-TM II CVs use slightly different section headings depending on whether the contributor is an employee or an external service provider, and the prompt fields can be written to treat both as equivalent.

**Does this kind of profile translate the CV?**

Only if you configure it to, and typically only for job title and project name, when the CV’s working language requires it. The summary, project description, and roles and responsibilities are best left untranslated and unreworded.

**Can the same approach be used for other framework contracts?**

Yes. This is one example built on AI Prompt Profiles, the same underlying capability. The same section-mapping logic can be configured as its own profile for any other recurring CV template.

## Final Thought

A framework contract migration is not the moment to introduce guesswork into a consultant’s CV. A profile built this way runs on the opposite instinct: extract only what is there, place it exactly where it belongs, and leave a field empty rather than fill it with something that was never written.

It is not a separate DIGIT-TM tool, just AI Prompt Profiles configured for one migration many consulting and staffing firms are living through right now. The same approach is available for whatever fixed CV format your team handles next.

Onboarding consultants under DIGIT-TM III and want this conversion set up for your own team? [Book a demo with Marco Pincho](https://calendly.com/marco-pincho/sprintcv/) to walk through your specific case.
