
How do I create a cosmetic PIF?
Five steps from your documents to a compiled dossier, and which parts the platform performs for you.
To compile a cosmetic Product Information File (PIF), you assemble the documentation for the product, screen every substance against the applicable restrictions at its use concentration, have the formulation assessed for safety, and consolidate the result into a single dossier meeting Article 11 of Regulation (EC) No 1223/2009. Performed manually, that means collating files, cross-checking against the Annexes and drafting the report. With dermaGPT you upload what you already hold, and the platform classifies the documents, identifies the gaps, screens the composition and generates the report. The safety assessment itself is reviewed and signed by an assessor qualified under Article 10(2).
What you need before you start
Nothing has to be organised in advance. Collect the documentation you already hold for the product, in PDF. Scanned paper documents are acceptable. What matters most:
- The qualitative and quantitative composition.Every substance and its percentage in the formulation. This is the one item without which the dossier cannot be built — exposure cannot be calculated, and no Annex restriction can be applied, without concentrations.
- Product identification.Name, intended function, and physical form — emulsion, gel, aerosol and so on — since the applicable restrictions depend on product type.
- The labelling and an image of the packaging.Both are required before that section can be treated as complete; the packaging material also bears on the Part A assessment.
- Physicochemical and quality data.Safety data sheets and certificates of analysis for the finished product and for the raw materials, obtained from your suppliers. These carry the purity and impurity information Annex I requires.
- Manufacturing, claims and testing records.Whatever you hold on how the product is made and on the evidence supporting each claim.
Gaps are normal at this stage. The platform identifies what remains outstanding once the upload is complete.
Compiling a PIF in five steps with dermaGPT
- Upload what you hold.Add your PDFs individually, or upload an entire folder: the platform extracts the PDFs, skips other file types and reports how many were skipped. If you already hold a completed PIF report, upload that instead and it is separated into its constituent chapters.
- Let it classify and check.Each file is assigned to the appropriate section of the dossier. A checklist shows which required sections are complete and which are outstanding, and identical files are flagged for removal in one action. Any misclassified file can be recategorised from a dropdown.
- Confirm identification and composition.Enter the product name, category and form. A range with several shades or fragrance variants can be entered as sub-products. The ingredient table is extracted from your documents, so you verify it rather than transcribe it. Before confirmation, ingredient names are checked for nomenclature errors and corrections are proposed; you decide whether to accept them. Accurate INCI naming matters here, because a misspelled name will not match the Annex entry that governs it.
- Run the analysis.One action starts it. The platform validates the dataset, performs the exposure and margin calculations, screens each substance against the applicable restrictions for your target markets and generates the report. This takes a few minutes; you may leave the page and are notified by email and in the application on completion.
- Review, sign and retain.A qualified safety assessor reviews the evidence and signs Part B of the safety report. The signed report, its version history and the source documents are held together in one library, available to a competent authority on request at the address of the Responsible Person.
What dermaGPT does for you
The laborious and error-prone parts of a PIF are the collation, the screening and the assembly. Those are what the platform automates.
| Task | Manually | With dermaGPT |
|---|---|---|
| Collating documents | Rename, file and track documents across folders and email threads | Files are classified into the required sections automatically, and duplicates are flagged |
| Identifying gaps | Compare your holdings against a checklist yourself | A live checklist shows complete and outstanding sections, and the analysis will not run without the mandatory ones |
| The composition table | Transcribe it from supplier documentation | Extracted from your PDFs into a table, with a nomenclature check on ingredient names |
| Substance restrictions | Look up each substance in each market's lists, at its concentration and for the product type | Screened automatically against the markets you supply |
| Generating the report | Assemble the dossier section by section | Generated from your dataset in a single run |
| Safety sign-off | Locate an assessor and exchange files back and forth | Reviewed and signed by a qualified safety assessor |
| Keeping it current | Multiple versions across multiple folders | One library holding every report, version and source document |
Common findings
- Starting without complete concentrations.A full quantitative composition is the one input the report cannot be produced without. An INCI list alone is not a formula.
- Supplying one half of a pair.Labelling without a packaging image, or a product data sheet without the corresponding raw material certificates, leaves that section outstanding.
- Ignoring the outstanding list.The checklist states precisely what is missing. Approaching suppliers early saves the most time, since raw material documentation is usually the longest lead item.
- Skipping the nomenclature check.Correct INCI names are what allow a restriction to be matched to the right substance; a misspelling can mask a restricted ingredient entirely.
- Treating the draft as final.The report is complete only once a qualified safety assessor has reviewed and signed Part B. Until then the product may not be placed on the market.
- Failing to issue a new version.A change of formulation, supplier, specification or claim, or an amendment to the Annexes, requires the file to be updated.
Frequently asked questions
- What do I need to upload to create a PIF?
- The documentation you already hold for the product, as PDFs: the quantitative composition, product identification, the labelling and a packaging image, and supplier data sheets and certificates. The platform reports what remains outstanding.
- Do I have to organise my files first?
- No. Upload them as they are, or an entire folder at once. Each file is classified into the appropriate section, and you can correct any classification.
- I already hold a PIF. Can I upload it?
- Yes. Upload the completed report and it is separated into its chapters, so you can maintain it from there.
- What file types can I upload?
- PDF only. Paper documents can be scanned to PDF. When a folder is uploaded, other file types are skipped and the count is reported.
- How long does it take?
- Uploading and review depend on the volume of documentation. The analysis itself runs for a few minutes in the background, with notification on completion.
- Who signs the safety report?
- A qualified safety assessor, as required by Article 10(2). The platform compiles the data and drafts the report; a qualified person reviews the reasoning and signs it.
- Does dermaGPT file the PIF or notify the product for me?
- No. It compiles and maintains the file. You or your Responsible Person remain responsible for any notification, such as the CPNP notification required in the EU before placing on the market.
- Can I create one PIF for a whole product range?
- Yes. Where a range comprises several shades or fragrance variants, each is entered as a sub-product with its own name, function and form. Note that variants differing in composition still require their own assessment of the differences.