| Account creation and login | Controller | Users of the platform (employers, HR staff, candidates, authorised representatives) | - Basic identifiers such as name, email, phone and company information
- Authentication data like hashed passwords, multi-factor authentication (MFA) secrets and refresh tokens
- Security-related data including session IDs, truncated IP addresses, browser user-agent strings and login timestamps
- User preferences and profile settings
| - Art. 6(1)(b) — Contract: Required to create and manage user accounts ('necessary for the performance of a contract').
- Art. 6(1)(f) — Legitimate interest: Collecting technical and security data strictly to protect the service from misuse and ensure network and information security (Rec. 49).
- Art. 32 — Security of processing: Obligation to implement state-of-the-art technical and organisational measures to safeguard accounts.
| - We provide a short, clear and easily accessible privacy notice at sign-up (Art.13) so users know exactly what data is collected and why.
- We apply data minimisation (Art. 5(1)(c)): only the information essential for creating and maintaining an account is required.
- Passwords are hashed and never stored in plain text.
- To limit risk, inactive accounts are automatically deleted after 12 months. Before deletion, users are given a 30-day window to reactivate if needed.
|
| Identity and background verification (Permly Verification) | Processor (the employer or applicant is the Data Controller) | Applicants, Family members (if included in the application) | - Mandatory identification documents required by Migrationsverket, such as passport scans (including the machine-readable zone, MRZ) and, where applicable, national ID cards or travel documents.
- Mandatory biometric data, such as a facial image (passport-style photo) for identity verification, and other biometric information.
- Mandatory qualifications, including education certificates, diplomas, CV/resumé, and any other supporting evidence of professional experience needed for the permit type.
- Mandatory contact details.
- Mandatory personal and family information, such as marital status, family composition, and other civil status details.
- Optional supporting documents voluntarily provided by the applicant (e.g., additional references, proof of experience, or other records for their own case-handling or documentation purposes).
| - Art. 6(1)(b) — Contract: Processing is necessary to perform pre-contractual and contractual steps for employment or relocation.
- Art. 6(1)(c) — Legal obligation: Required for compliance with the Swedish Aliens Ordinance (Utlänningsförordningen, Ch. 4) and related migration regulations.
- Art. 9(2)(b) or (f) — Special-category data: Processing sensitive data (e.g., biometric data) strictly where required by law or necessary for establishing, exercising or defending legal claims.
- Art. 28 — Processor obligations: Processing is performed strictly under a Data Processing Agreement with the Controller, in accordance with GDPR Art. 28.
| - Processing limited to the minimum data explicitly required by Swedish migration law; any additional data stored is voluntary and user-controlled.
- A signed Data Processing Agreement (DPA) with all required Art. 28(3) clauses, defining scope, purpose and security obligations.
- Encryption in line with industry best practice.
- Transparent sub-processor management: maintaining and publishing an updated list of all sub-processors with at least 30-day advance notice before changes (Art. 28(2)).
|
| Power-of-Attorney and document e-signing | Processor | Applicants, Employers | - Signed Power of Attorney (POA) in PDF/A-3 with embedded Qualified Electronic Signature (QES)
- Application forms and supporting documents in PDF/A format
| - Art. 6(1)(c) — Legal obligation: Required for compliant submission under Förvaltningslagen §14.
- Art. 28 — Sub-processor engagement: Use of an EU/EES-based Qualified Trust Service Provider (QTSP) for advanced/qualified electronic signatures, with prior notice to the Controller.
| - Execution of Qualified Electronic Signatures (QES) via an EU-listed QTSP with BankID or equivalent strong identity proofing.
- Immutable, hash-chained audit trail with RFC 3161-compliant time-stamps for full traceability.
- Dual confirmation: both employer and applicant must review and approve documents before submission.
|
| Employer compliance & job offer assembly (Migrationsverket Form 232011) | Processor | Employers, Applicants | - Job details: including but not limited to title, core duties (aligned with SCB/SSYK taxonomy), SSYK 4-digit code, employment type (permanent/fixed-term), scope of work (full-time/part-time), and contract duration.
- Employment terms: full employment agreement text including annexes, probation clauses, collective bargaining references (if applicable), union contact details, and termination conditions.
- Compensation: base salary, variable remuneration (bonuses/commissions), benefits (e.g., meal, housing, travel allowances), annual salary review process, and working time details as required under MIGRFS 2023:1.
- Insurance coverage: proof of mandatory insurances including occupational injury (TFA), health, life, and occupational pension (tjänstepensionsförsäkring), including policy numbers and providers, consistent with Migrationsverket's requirements for work permit eligibility.
- Workplace information: employer organisation details (Bolagsverket registration number), physical work location, organisational unit, and designated company contact person.
- Leave entitlements, if applicable: annual leave days (per Semesterlagen), parental or family leave entitlements, and any additional paid/unpaid leave policies.
| - Art. 6(1)(b) GDPR — Contract: Processing is necessary for the preparation of an enforceable employment agreement between employer and applicant, a prerequisite for work permit eligibility under Swedish law.
- Art. 6(1)(c) GDPR — Legal obligation: Required for compliance with Chapter 4 of the Swedish Aliens Ordinance (Utlänningsförordningen) and the Swedish Migration Agency's regulation MIGRFS 2023:1 on salary, insurance, and employment conditions.
- Art. 5(1)(c) GDPR — Data minimisation: Only the information explicitly required by Migrationsverket's Form 232011 and associated legal frameworks is collected; optional information is clearly marked as voluntary.
- Art. 28 GDPR — Processor obligations: Processing performed strictly under a Data Processing Agreement (DPA) with the employer (Controller), with all mandatory Article 28(3) provisions (purpose, instructions, confidentiality, sub-processing, and technical measures).
- Art. 5(2) GDPR — Accountability: Change-tracking and review workflows allow the Controller to demonstrate compliance in case of audits or Migration Agency inquiries.
| - Structured, schema-driven digital interface that replicates Migrationsverket's official Form 232011 layout, enforcing mandatory fields and data formats to prevent omissions or errors.
- Automated SSYK classification engine. Employers can override suggestions to preserve Controller discretion (aligning with GDPR Art. 24).
- Real-time compliance validation: salary, benefits, and insurance coverage cross-checked against MIGRFS 2023:1 thresholds and Statistics Sweden (SCB) benchmark data, with API-based timestamped lookups to evidence the regulatory alignment at the time of assembly.
- Visual traffic-light system (green/yellow/red) flags incomplete or non-compliant data, reducing the likelihood of rejection by Migrationsverket.
- Immutable audit log: cryptographically linked records capture every modification (who/what/when), satisfying GDPR Art. 5(2) and providing evidentiary value in disputes or inspections.
- Dual-review process: the employer must explicitly confirm the job offer's accuracy, and the applicant is presented with a read-only version for review and consent prior to submission.
- Exported job offers stored as PDF/A-3 with embedded SHA-256 hash and RFC 3161-compliant trusted timestamps, ensuring document integrity and non-repudiation.
- Role-based access controls (RBAC) and least-privilege principles ensure that only authorised employer representatives and designated Permly processors can access or edit the job offer, in line with GDPR Art. 32.
- Retention governed by Bokföringslagen (1999:1078) and migration law: job offer data retained only for as long as legally necessary for case handling or regulatory compliance, then securely deleted or anonymised.
|
| Secure communication with authorities | Processor (for transmission) / Independent Controller (for limited legal-defense retention) | Applicants, Employers, Family members | - Case identifiers (diarienummer, case IDs)
- Application forms and supporting documents (PDF/A)
- Audit metadata (submission timestamps, sender identity, authority receipt confirmations)
| - Art. 6(1)(c) GDPR — Legal obligation: Required for compliant submission of applications to Migrationsverket and other competent authorities.
- Art. 6(1)(f) GDPR — Legitimate interest: Retention of minimal proof of transmission (hash, timestamp, receipt) for up to 12 months to defend against service disputes or authority inquiries.
- Art. 13(3) GDPR — Transparency: Informing data subjects when the competent authority assumes Controller responsibility upon receipt of the transmitted data.
- Art. 28 GDPR — Processor obligations: Processing performed under Controller instructions pursuant to a Data Processing Agreement (DPA).
| - Data is transmitted using PEPPOL/SFTI channels or direct authority APIs with mutual TLS and RFC 5280-compliant certificate validation.
- Two-tier retention model: (a) operational copies retained in encrypted form for up to 12 months solely to handle appeals, authority queries, or dispute resolution, then securely deleted or anonymised (GDPR Art. 5(1)(e)); (b) where instructed by the employer (Controller), archival copies of employment-related documents qualifying as accounting records (e.g., job offers) retained in a restricted-access encrypted archive for up to 7 years in compliance with Bokföringslagen (1999:1078, Ch. 7 § 2).
- All transmissions logged in a hash-linked, immutable audit trail with RFC 3161-compliant timestamps, enabling full traceability for Controllers and authorities (GDPR Art. 5(2)).
- Users notified when their data is handed over to an authority, with clear information on the authority's role as new Controller (GDPR Art. 13(3)).
- Role-based access (RBAC) and least-privilege principles applied to all transmitted data; archives stored with AES-256-GCM encryption in EU/EES data centres.
|
| Platform analytics, cookies, marketing communications, and voluntary feedback | Controller | Platform users, Website visitors, Prospects, Subscribers, Respondents | - Truncated IP addresses, device IDs, clickstream, and pseudonymised analytics data
- Error logs for diagnostics and security
- Email addresses and names for marketing and survey participants
- Engagement data (e.g., opens, clicks) for newsletters
- Voluntary survey responses and optional contact details
- Consent status, version history, and opt-in/opt-out signals (DNT, GPC)
| - Art. 6(1)(f) GDPR — Legitimate interest for essential analytics and diagnostics (Recital 49)
- Art. 6(1)(a) GDPR — Consent for marketing communications, cookies, and voluntary surveys
- ePrivacy Directive 2002/58 & Swedish Marketing Act 2008:486 — Consent for tracking and marketing
| - Documented Legitimate Interest Assessment (LIA) for analytics (Rec. 49).
- Use of a Consent Management Platform (CMP) with granular, IAB TCF v2.2-compliant preferences; consent text, timestamps, and proof retained for 5 years.
- Truncation and pseudonymisation of analytics data; differential privacy applied for aggregated insights.
- One-click unsubscribe for all marketing communications; suppression lists maintained to honour opt-outs.
- Respect for browser signals (Do Not Track, Global Privacy Control).
- Retention: analytics logs deleted after 12 months; survey data anonymised or deleted after 24 months; consent records retained for 5 years.
|
| Aggregated benchmarking, statistics, and AI-assisted analytics | Out of GDPR scope after anonymisation | Controller-provided employer and applicant case data (processed under DPA until anonymisation), Aggregated, irreversibly anonymised datasets across multiple employers (no personal data) | - Case metadata (e.g., intake and decision dates, SSYK codes at 2-digit level, employer size buckets, and processing durations), stripped of all direct identifiers.
- Aggregated statistical measures (e.g., average processing times, workload bottlenecks, approval rates) across multiple customers, k-anonymised and suppression-adjusted.
- No linkage keys, free-text data, or uniquely identifying combinations are retained in anonymised outputs.
| - Art. 28 GDPR — Processor obligations: All raw and pseudonymised controller data is processed strictly under the employer's instructions in accordance with our Data Processing Agreement (DPA).
- Recital 26 GDPR — Anonymous data: Once datasets are irreversibly anonymised using generalisation, suppression, and k-anonymity, they are no longer considered personal data: 'The principles of data protection should not apply to anonymous information, namely … data rendered anonymous in such a manner that the data subject is not or no longer identifiable.'
- Art. 6(1)(f) GDPR — Legitimate interest: processing minimal, non-sensitive, pseudonymised metadata for the sole purpose of service improvement (e.g., identifying process inefficiencies), supported by a documented balancing test (Recitals 47 & 49).
- Art. 35 GDPR — Data Protection Impact Assessment (DPIA): DPIAs are conducted for analytics workflows involving AI, covering re-identification risk, proportionality, and bias mitigation.
- EU AI Act — Limited-risk classification: Internal AI-assisted analytics is categorised as 'limited-risk' under Title III, Chapter 1, as no automated decisions with legal or significant effects are made (Recital 60).
| - Anonymisation process: Apply a documented anonymisation SOP including quasi-identifier generalisation (e.g., SSYK reduced to 2-digit groups, employer size bucketed as <50 / 50-249 / 250+), date shifting to month/quarter granularity, k-anonymity (k ≥ 10), and suppression of small cohorts. Free-text fields are mapped to controlled taxonomies or excluded entirely. Linkage keys are deleted post-processing. (See WP29 Opinion 05/2014.)
- Strict segregation: Until anonymisation is complete, all working datasets segregated with role-based access control (RBAC), encryption at rest, and encrypted linkage keys accessible only to authorised personnel.
- Bias & fairness testing: All internal machine-learning components used for process analytics undergo bias and performance testing at deployment and quarterly thereafter, following EU AI Act Art. 10 and ISO/IEC 24027:2021 guidelines.
- Model documentation: Maintain model cards, version control, and change logs for all AI components, aligning with transparency principles in the EU AI Act and ENISA best practices.
- User rights: Pre-anonymisation, data subjects retain their GDPR rights (Art. 15-21), including the right to object (Art. 21) to inclusion in benchmarking datasets. Post-anonymisation, outputs are fully anonymous and fall outside GDPR's scope.
|
| Rights handling & breach notification | Controller | Any data subject, Potentially affected individuals in case of a breach | - Personal data necessary to verify and respond to rights requests (access, rectification, erasure, portability)
- Incident data: categories of affected data, risk assessment, and summary for notification purposes
| - Art. 6(1)(c) — Legal obligation (GDPR Articles 12-23, 32-34)
| - Rights requests handled via secure portal with ID verification
- Responses issued within 30 days (extendable +60 days)
- Machine-readable export (JSON/CSV) for portability
- Incident detection, risk assessment, and root-cause analysis
- Notify IMY within 72 hours for reportable breaches (Art. 33) and inform individuals when risk is high (Art. 34)
- Audit logs maintained for all rights requests and incidents
|