Data Processing Agreement (DPA) — Mondigo (United Kingdom)

Version: mondigo-gb-dpa-2026-11-16-v1.2 · In force from: 2026-11-16 (previous version: mondigo-gb-dpa-2026-10-31-v1.1, 2026-10-31) · Based on the Hungarian Adatfeldolgozási megállapodás mondigo-dpa-2026-11-16-v1.6-price-tip-removed

In force from 16 November 2026. This English agreement is the binding data processing agreement for Mondigo workspaces registered for the United Kingdom; it is not a translation provided for information only.

v1.2 (2026-11-16): sections 3.6 and 3.7 and the operations row of Annex A no longer refer to the "save as own price" offer, which has been removed from the service: the sentence that an own price is created from an offer based on a learning pattern only by the tenant's explicit decision is deleted; "related offer data" becomes "related price-suggestion data"; "forming patterns within the tenant and offering them" becomes "forming patterns within the tenant" (as in the Hungarian DPA v1.6). No new subprocessor and no new data category. Wording approved by the owner 2026-11-16 10:34Z; the DPA version change awaits an owner card.

v1.1 (2026-10-31): at the end of the service the tenant chooses between the return and the deletion of the personal data (section 12.2); the audit and inspection right under Article 28(3)(h) of the GDPR and of the UK GDPR is stated expressly (section 13). No new subprocessor.

v1.0 (2026-10-30): first version for UK workspaces, prepared from the Hungarian DPA v1.5 (sections 3.5–3.11 as in v1.5); UK GDPR references and the transfer wording of section 9 added for UK users. No new subprocessor.

The User accepts this agreement electronically at registration by ticking the checkbox (Article 28(9) of the GDPR and of the UK GDPR); a separately signed copy is not required (see section 15).

In this agreement “GDPR” means Regulation (EU) 2016/679, and “UK GDPR” means the GDPR as it forms part of the law of the United Kingdom, read with the Data Protection Act 2018. Where this agreement refers to an Article of the GDPR, the corresponding Article of the UK GDPR also applies to the extent the UK GDPR governs the processing.

1. Parties

Processor / Provider:
ZST Radio Korlátolt Felelősségű Társaság (ZST Radio Kft.)
Registered office: 1037 Budapest, Kisbojtár utca 35. fszt. 2., Hungary
Company registration number: 01-09-375210
Hungarian tax number: 28814012-2-41
Contact: info@zstradio.com

Controller / tenant (User):
The User who registers for the Mondigo service for a United Kingdom workspace, with the data given at registration (or when accepting an invite): name or business name, email address, and — if given — VAT registration number, Companies House number, phone number and service area. The registration data identifies the Controller in the contract between the Parties; the Controller's contact person can be reached at the registered email address.

Hereinafter together: the Parties.

2. Subject matter and relation to the main contract

2.1. This DPA governs the terms of the processing of personal data connected with the use of the Mondigo SaaS service, for the data flows in which the tenant determines the purposes and essential means of the processing of personal data and ZST Radio Kft. acts on the tenant's documented instructions.

2.2. This DPA supplements the Mondigo Terms of Service (United Kingdom). In case of conflict, on questions of the processing of personal data, mandatory data protection law and then the provisions of this DPA prevail.

2.3. This DPA does not apply to the purposes for which ZST Radio Kft. acts as an independent controller, for example its own account administration, service security or legal-obligation purposes; these are covered by the Mondigo Privacy Notice (United Kingdom).

3. Subject matter, duration, nature and purpose of the processing

3.1. Subject matter: the technical processing, through the functions of Mondigo, of the personal data needed for the tenant's customers, jobs, quotes and related organisational and communication processes.

3.2. Duration: the period during which the tenant uses the Mondigo service, and the deletion/return period that follows it under section 12.

3.3. Nature of the processing: collection from the tenant's input, structuring, storage, retrieval, alteration, quote generation, AI-assisted interpretation, quality-assurance recording and regression replay of AI suggestions, email delivery, customer-facing display, recording of statuses and events, export and deletion.

3.4. Purpose: supporting the tenant's own business activity, in particular preparing and sending quotes, customer feedback, arranging appointments and organising work, including the quality assurance of the AI suggestions that support quote preparation.

3.5. Quality-assurance recording and regression replay: ZST Radio may record, for each quote, the evidence of the structuring (the note with the transcript, the raw model reply and the line items, the result displayed, the decisions of the decision model, the edits of the tenant's user, the items and amounts of the sent version, the customer's response without the free-text reason for a rejection, the model/release/timing; according to a fixed list of fields), exclusively to measure the quality of AI suggestions, to replay earlier model replies through newer versions of the processing as a regression check, and to analyse errors. The end customer's structured contact details (name, email, phone, address) are not included in the record, and the contact details recognised are removed from the texts kept (the filtering is not complete). The records stay in the EU (Microsoft Azure, Poland Central), with the processors listed in Annex B; they may be kept for 90 days from recording (longer only under a documented legal hold or during an active incident; a copy deleted from the storage can be restored for at most 7 more days; the backups under section 12 do not contain them); they are deleted with the tenant, with the quote and, on the tenant's instruction, per data subject; they are included in the tenant's data export; at the tenant's request they can be excluded from recording and from transfer to the storage (with deletion of the existing records); regression replay — once available — runs only in Azure's EU environment; and they may not be used for model training or fine-tuning.

3.6. Loading earlier quotes

The tenant may give Mondigo a CSV or XLSX data file of its own earlier quotes, or — if the function is active — a document in PDF, JPG or PNG format, so that Mondigo prepares own price and usual job suggestions for the tenant.

For this processing the tenant is the controller and ZST Radio Kft. the processor. The data subjects are primarily the tenant's former customers and contact persons.

The whole file is held in the memory of Mondigo's server only temporarily, for the time of the processing. Only the data needed to process the quote item may be stored durably. The raw file and its file name are not stored durably.

The processing takes place in the Microsoft Azure Poland Central environment. Mondigo's server processes a CSV/XLSX file without AI. Mondigo reads a PDF, JPG or PNG document with an AI model on the Microsoft Azure AI Foundry EU Data Zone route listed in Annex B: a PDF is sent to the model unchanged, an image after metadata removal, face blurring and downsizing where needed; at most 5 documents per load, each at most 8 MB. Following a fixed instruction, the model returns only item data (description, quantity, unit, unit price, price type, VAT rate, date, trade) and no data relating to a person; the same data minimisation applies to the descriptions returned as to CSV/XLSX items. The document, the image sent to the model and the model's raw reply are not stored durably, are not included in the recording under section 3.5, and may not be used for model training or fine-tuning. No new processor is engaged for this.

The loaded item rows (from a CSV/XLSX file and from a document alike) are kept for at most 90 days from loading, or for a shorter time if the tenant deletes the import or its account. The production backups do not contain the loaded item rows.

The tenant instructs Mondigo to minimise personal data not needed for the purpose of the service, and not to use item descriptions recognised as a personal name or an address for a lasting own price, usual job or learning pattern. Before loading, the tenant removes — for a document, where possible covers — the personal data not needed to describe the work item.

3.7. Pricing patterns within the tenant

From the item prices entered or changed by the tenant in the quotes it sent, Mondigo may form pricing patterns usable exclusively for that tenant.

The purpose of the processing is to recognise the tenant's own consistently applied prices. A price that the tenant entered or changed for a quote item goes into the tenant's own price list; an item description recognised as a personal name or an address does not go into the own price list.

A learning pattern and the learning evidence belonging to it may not contain a customer name, customer address or customer contact details. Customer data may be used only temporarily, to determine with how many different customers the same work item occurred. An item description recognised as a personal name or an address may not be used to form a learning pattern.

The learning evidence may refer to the technical identifier of the underlying quote. When a quote is deleted, the learning evidence linked to it must also be deleted, and the affected pattern must be recalculated from the remaining evidence or deleted. The tenant cannot delete a sent quote; the learning evidence of such a quote is deleted when the tenant's account is deleted or when a data subject's erasure request is fulfilled.

The patterns and the related price-suggestion data may not be used for another tenant, may not be used to train or fine-tune general models, and may not be passed on to further providers.

The patterns may be kept at most until the tenant's account is deleted.

3.8. On-site photos in the quote

If the function is active, the tenant (the tradesperson) may attach on-site photos to its quote. ZST Radio processes the photos on the tenant's instruction, exclusively for preparing the quote, displaying it to the tenant's customer, exporting and deleting it:

a) Processing on upload: removal of the photo's metadata (EXIF, GPS, XMP, ICC), applying the orientation, blurring the faces detected, downsizing in two variants. A photo without blurring is not stored. Face detection is not complete; it is the tenant's task to upload only images needed for the work.

b) Storage: Microsoft Azure Blob Storage, Poland Central (EU), a dedicated storage account and private container, application-level AES-256-GCM encryption with a separate data key for each photo, wrapped by a separately managed master key of the Provider; access exclusively with the managed identity of the Provider's server, without shared keys or public access.

c) Display to the tenant's customer: only the photos approved and left visible by the tenant before sending, as part of the fixed (hash-protected) content of the sent quote, on the public quote link; with headers that forbid caching, search engine indexing and forwarding of the referrer header. The customer does not receive a hidden photo. The email does not contain photos.

d) AI (with a separate switch): a photo may go to the Microsoft Azure AI Foundry EU Data Zone route (Annex B) only to assign it to an item and to suggest a caption; no quantity, price or other quote data is determined from the photo; the photo may not be used for model training or fine-tuning. The tenant checks the suggestion.

e) Retention and deletion: a photo is kept together with the quote; the photo of a sent quote cannot be changed. A draft containing only photos (no items and no note) is deleted automatically, together with its photos, 14 days after its last change. When a photo, the draft or the tenant is deleted, the stored image files are deleted too. A deleted file can be restored in the store for at most 90 days (soft delete and versioning), after which it is permanently deleted. In case of a restore, deletions completed earlier must be applied again under section 12.

f) Export: the metadata of the photos is included in the tenant's data export; the photos can be downloaded one by one from the tenant's account.

g) The tenant's responsibility: the tenant as controller is responsible for the content of the photos, for the lawfulness of processing the personal data shown in them (including an image of the customer's home or of a third person), and for which photos it makes visible to the customer.

3.9. Post-job customer rating

If the function is active, by marking the job as done the tenant instructs Mondigo to send, on the tenant's behalf, a single rating request by email to the tenant's customer, with a link tied to the job and valid for 60 days from completion. Mondigo stores and displays the answer (the recommendation answer, four ratings from 1 to 5, how the job turned out compared with the quote, whether the change was agreed in advance, a written comment of at most 2,000 characters, timestamps, the number of signed variations recorded for the quote and the change in their amount) exclusively for the tenant. It does not make the rating public, does not make it available to another tenant and does not pass it to AI processing; only content-free, pseudonymous process events go to product analytics. Retention: together with the quote, until the tenant's account is deleted; it is part of the tenant's data export. No new processor.

3.10. Operator console

To manage Mondigo accounts (invites, access, free periods, customer support), ZST Radio uses an internal operator console that only its designated staff, bound by confidentiality, can access, with two-factor authentication. The console shows account-level data about the tenant (the account's email address, company name, trade, dates, counts, usage counters, access status) and — with a separate switch — the pseudonymous product analytics events of the tenant or of individual users of the tenant (in the user-level view with the user's email address and role). It does not display the personal data of the tenant's customers (name, contact details, address), the content of quotes and items, the notes or the photos. These data fall within ZST Radio's independent controllership; ZST Radio continues to access the tenant's customer data only under section 6, to the extent needed to provide the service, for troubleshooting or to comply with a legal obligation. ZST Radio logs operator actions.

3.11. Recognition of usual jobs and change suggestions

For the tenant's saved usual jobs, Mondigo — according to the functions the tenant has switched on — carries out the following processing, exclusively within the tenant's account, without AI and without a new processor: (a) if the note contains, word for word, the name of a usual job or the nickname given to it by the tenant, it inserts the items of that usual job into the quote draft by a fixed rule, before the AI processing, and records which version of which usual job went into the quote, how and with what multiplier; (b) when the quote is sent, it records the item list of the usual job used at that time, and the description, unit and quantity of the items of the sent quote (not the prices of the sent quote or the customer's contact details), and from these it may suggest to the tenant's owner a change to the default quantity or quantity mode of the usual job. The usual job changes only with the owner's explicit acceptance; the sent quote does not change. Retention: until the tenant's account is deleted (usage data recorded for a draft, with the deletion of the draft); part of the tenant's data export; not usable for another tenant or for model training.

4. Data subjects and categories of data

4.1. Categories of data subjects

4.2. Categories of personal data

4.3. The tenant does not instruct Mondigo to process, as a matter of routine, special categories of data (Article 9 GDPR / UK GDPR), criminal offence data or sensitive data not needed for the purpose of the service. If such data nevertheless gets into free text or an image, the tenant must minimise it and ensure its lawfulness.

5. Obligations of the tenant as controller

The tenant:

6. Obligations of ZST Radio / Mondigo as processor

ZST Radio Kft.:

If, in ZST Radio's opinion, an instruction of the tenant infringes the applicable data protection law, it informs the tenant without undue delay and may restrict the risky processing until the lawful position is clarified.

7. Technical and organisational measures

The measures applied include in particular:

ZST Radio makes the detailed technical documentation available under section 13, at the tenant's reasonable request.

8. Further processors (subprocessors)

8.1. The tenant gives a general written authorisation to engage the further processors on the current list published in section 8 of the Privacy Notice (United Kingdom).

8.2. ZST Radio notifies the tenant at least 14 days in advance of the planned engagement of a material new or replacement subprocessor, except in an urgent security/legal case, when notification takes place without undue delay.

8.3. The tenant may object on well-founded data protection grounds. The Parties look for an alternative solution in good faith; if this is not reasonably possible, the tenant may stop using the service concerned.

8.4. ZST Radio imposes on its subprocessors data protection obligations essentially equivalent to those required by the applicable data protection law, and maintains the chain of responsibility under the GDPR and, where it applies, the UK GDPR.

9. International data transfers

9.1. The active email route of the minimum pilot is Amazon SES eu-central-1; the core hosting/backup and the AI route are also configured for a documented EU region. This technical choice of region does not replace the review of each provider's full contractual terms, account/support metadata and any applicable international data transfer.

9.2. Such a transfer may take place only on the basis of an applicable mechanism under Chapter V of the GDPR, for example an adequacy decision, the EU Standard Contractual Clauses or another valid safeguard.

9.3. The current list of processors and the configured regions are set out in the table published in section 8 of the Privacy Notice (United Kingdom). ZST Radio makes no general claim that “all data stays exclusively in the EU” in a configuration where that is not technically true.

9.4. The tenant, established in the United Kingdom, transfers the personal data it enters to ZST Radio in Hungary (EEA). Under the UK GDPR this transfer relies on the adequacy regulations by which the United Kingdom recognises the EEA countries (section 17A of the Data Protection Act 2018). Where a subprocessor processes UK personal data outside the United Kingdom and the EEA, the onward transfer may take place only on the basis of a mechanism under Chapter V of the UK GDPR, for example UK adequacy regulations (including the UK Extension to the EU–U.S. Data Privacy Framework for a certified recipient) or the UK International Data Transfer Addendum to the EU Standard Contractual Clauses.

10. Data subject requests

10.1. If ZST Radio receives a data subject request that clearly belongs to the tenant's processing of customer data, it forwards the request to the tenant — where lawful and possible — and does not take a controller's decision on the merits in its place.

10.2. Mondigo supports the fulfilment of requests for access, rectification, erasure, restriction, portability and other applicable requests with the technical functions available and reasonable manual support.

10.3. The operational target times and process for handling data subject requests are set out in ZST Radio's internal policy.

11. Personal data breaches

11.1. If ZST Radio becomes aware of a personal data breach affecting the data it processes for the tenant, it notifies the tenant without undue delay. The internal operational target is a first substantive notification within 24 hours, where sufficient information is available; this does not change the legal standard of the GDPR or the UK GDPR.

11.2. To the extent available, the notification contains the nature of the breach, the categories of data concerned, the likely consequences, the measures taken or planned and the contact point.

11.3. The tenant decides, in its capacity as controller, on its obligations to notify the supervisory authority (for a UK tenant, in particular the Information Commissioner's Office) and the data subjects; ZST Radio provides reasonable support for this.

12. Return, deletion and backup

12.1. While the service exists, the tenant may request the release of its own data with the available export function.

12.2. At the end of the provision of the services relating to the processing, the tenant as controller chooses whether ZST Radio returns the personal data to it (through the export function or another agreed, commonly used electronic format) or deletes it, and ZST Radio then deletes the existing copies, unless applicable law (United Kingdom law, the law of the European Union or of Hungary) requires their further storage, and then only to the extent and for the time it requires. Unless the tenant has asked for the return earlier, the tenant's business/customer data is deleted from the live production system within at most 30 days after the service ends; the data can be exported until the account is deleted.

12.3. In-place deletion of individual records from the production backups is not part of the normal process. Backups expire under a 90-day lifecycle; previous object version retention is 7 days; expired copies can then still be restored during a deletion-protection (soft delete) period of at most 30 days. In case of a restore, deletions completed earlier must be applied again before the restored data is put back into active use.

12.4. The tenant may request reasonable confirmation that the deletion has taken place.

13. Audit and compliance evidence

13.1. In line with Article 28(3)(h) of the GDPR and of the UK GDPR, ZST Radio makes available to the tenant all information necessary to demonstrate compliance with the obligations of Article 28 and of this DPA, including the relevant security, backup and provider documentation, and allows for and contributes to audits, including inspections, conducted by the tenant or by another auditor mandated by the tenant.

13.2. The tenant gives reasonable prior notice of an audit or inspection. An on-site or in-depth technical audit is carried out in a proportionate manner, at a time agreed in advance, under appropriate confidentiality obligations of the auditor, and under security conditions that protect the other customers' data, security secrets and business secrets; these conditions may not be used to make the audit right ineffective.

13.3. If acceptable independent audit/certification/provider evidence serves the same purpose, it may be used first; this does not exclude an audit or inspection where that evidence is not sufficient to demonstrate compliance.

14. Liability and governing law

Liability, fees and governing law are governed jointly by the main contract (the Terms of Service (United Kingdom), under Hungarian law) and mandatory data protection law.

15. Entry into force and acceptance

DPA version: mondigo-gb-dpa-2026-11-16-v1.2 (in force from 16 November 2026; replaces mondigo-gb-dpa-2026-10-31-v1.1, in force from 31 October 2026)

How the tenant accepts: electronic acceptance at registration (Article 28(9) of the GDPR and of the UK GDPR). At registration the User accepts the Terms of Service, the Privacy Notice and this Data Processing Agreement by ticking one checkbox; this agreement is concluded between the Parties as part of the Terms of Service. A separately signed copy is not required; on request the Provider confirms the text of the agreement and the fact of acceptance.

Date of acceptance: the time of registration (or of accepting the invite). The Provider records the time of acceptance and the document versions accepted (Terms of Service, Privacy Notice, this agreement).

Annex A — details of the processing

FieldDescription
Subject matterTechnical processing of the Mondigo SaaS customer/job/quote workflow; processing of the tenant's earlier quotes (CSV/XLSX, and PDF/JPG/PNG read by AI) and forming pricing patterns within the tenant (sections 3.6, 3.7); recognition of usual jobs and change suggestions (section 3.11)
DurationDuration of the service + deletion/backup expiry period; on-site photos: the periods under section 3.8 e) (together with the quote; a draft with only photos is deleted after 14 days; a deleted image file can be restored for at most 90 days)
Data subjectsThe tenant's customers, contact persons, staff recorded by the tenant; the tenant's former customers and contact persons
DataContact data, work address, note, quote, event, communication, on-site photos (without metadata, faces blurred), caption, assignment, visibility (section 3.8); the necessary item data of earlier quotes, and the normalised work key formed from the item, unit, aggregated price data, occurrence counts and the technical identifier of the underlying quote; post-job customer rating (section 3.9); the PDF/image document of an earlier quote for the time of reading (section 3.6); use of usual jobs: item list version, the description, unit and quantity of the items of the sent quote, the owner's decisions (section 3.11)
Special categories of dataNot targeted; routine input forbidden/to be minimised
OperationsInput, storage, structuring, AI processing, quality-assurance recording and regression replay (section 3.5), sending, display, export, deletion; CSV/XLSX processing, data minimisation, forming price/job suggestions, forming patterns within the tenant (sections 3.6, 3.7); photo processing (metadata removal, face blurring), encrypted storage, display on the public link after the tenant's approval, 14-day draft deletion (section 3.8); sending the rating request, storing the answer and displaying it to the tenant (section 3.9); AI reading of PDF/JPG/PNG documents (section 3.6); recognition of usual jobs, recording of usage data and change suggestions (section 3.11)
PurposeSupporting the tenant's quote and work processes, including the quality assurance of AI suggestions; supporting the tenant's own quote preparation with suggestions based on its own earlier and own sent quotes

Annex B — subprocessors

The current list is the table published in section 8 of the Privacy Notice (United Kingdom).

No new processor for section 3.8. The function of the Microsoft Azure row is extended with: “encrypted storage of on-site photos (Poland Central)”; the function of the Microsoft / Azure AI Foundry row: “assigning photos to items and suggesting captions, if the photo AI function is active”. No new processor for the PDF/JPG/PNG branch of section 3.6 or for section 3.11 either; the function of the Microsoft / Azure AI Foundry row is extended with: “reading earlier quotes uploaded as a PDF or image, if the function is active”.

Annex C — TOMs

Section 7 of this DPA; ZST Radio makes the detailed security, backup and monitoring documentation available on request under section 13.