Back to start

Privacy Policy

Last updated: November 10, 2026

1. Data Controller

Teclead Ventures GmbH
Köpenicker Straße 126, Aufgang 1
10179 Berlin, Germany

Managing Directors: Malte Herberg, Tony Schumacher
Phone: +49 30 221 836 13
Email: info@teclead-ventures.de

2. Data Protection Officer

Manuel Schierenberg
c/o Teclead Ventures GmbH
Köpenicker Straße 126, Aufgang 1, 10179 Berlin, Germany
Email: manuel.schierenberg@teclead-ventures.de

3. Purpose and Scope of Processing

The Process Platform is a B2B application for structured capture, assessment and downstream processing of our customers' business processes. We process personal data for the following purposes:

  • Providing user accounts (sign-up, login, profile)
  • Running process discovery in the AI-assisted chat, including storage of chat histories
  • Optional voice input in process discovery (live transcription of spoken input — see § 11 for details)
  • Automated scoring of captured processes and generation of derived outputs
  • Communication with you regarding support and contractual matters
  • Operation, security and error analysis of the platform
  • Usage statistics on the MCP access (the interface through which AI clients such as Claude or ChatGPT work with your account): for the product team we count, per company and per user account, whether and how the access is used. No content is recorded, and the figures are not used for performance or behaviour monitoring (see § 4 and § 9)

4. Categories of Data Processed

  • Account data: name, business email address, company, role
  • Authentication data: password hash, session tokens
  • Content data: process descriptions you enter in the discovery chat, including any names of colleagues, customers or suppliers mentioned
  • Audio and transcript data (only when you actively start voice input): the live microphone signal and the partial / final transcripts derived from it (see § 11)
  • Speech output data (only when you actively switch voice mode on): the AI reply text sent to speech synthesis and the resulting audio, played in your browser only (see § 11)
  • Usage data: timestamps, pages visited, process interactions
  • Technical access data: IP address, browser type, operating system, referrer
  • MCP usage data (only when you use the MCP access): for each tool call, the name of the tool, the outcome (successful, refused, error, not authorised), the duration, the access path (OAuth or API token), with OAuth the identifier of the connected client, the time, your user account and company, when a process is saved or changed a reference (ID) to that process, and a user agent shortened to 100 characters. Arguments, process content, names mentioned in it and error message texts are not stored. Only platform administrators of Teclead Ventures GmbH can read this data, not administrators or consultants of your organisation.

Note on third-party data: If you mention personal data of third parties in the discovery chat (e.g. names of colleagues, clients or suppliers), responsibility for the lawfulness of this disclosure lies with your organisation as the data controller. Please ensure that your organisation has an appropriate legal basis for transmitting such data to our platform.

5. Legal Bases

  • Art. 6(1)(b) GDPR (contract performance): for providing the user account and the core platform features (discovery, scoring, output generation).
  • Art. 6(1)(f) GDPR (legitimate interest): Our legitimate interest lies in ensuring the operational security of the platform, detecting and resolving technical errors, and preventing misuse. This processing is necessary for these purposes and does not override your interests, as only technical data is affected and the processing serves to protect all users. For the usage statistics on the MCP access (§ 3) the legitimate interest is product development; only usage metadata without content is processed for it.
  • Art. 6(1)(c) GDPR (legal obligation): where tax or commercial retention duties apply.

6. Necessity of Providing Personal Data

Providing your email address and a password is contractually required for registration and use of the platform. Without this information, we cannot create a user account or provide the platform.

Providing your name, company and role during onboarding is voluntary, but enables a personalised experience and is recommended for full process discovery functionality.

Entering process descriptions in the discovery chat is required for the core functionality of the platform; you determine the nature and scope of the content you provide.

7. Recipients and Processors

We engage the following processors under Art. 28 GDPR to operate the platform:

  • Supabase Inc. (database & authentication) — 65 Chulia Street #38-02/03, OCBC Centre, Singapore 049513. Data is stored exclusively in the Stockholm region (EU, eu-north-1). Supabase uses Amazon Web Services (AWS) as a sub-processor in the same EU region.
  • Vercel Inc. (hosting & edge network) — 340 S Lemon Ave, Walnut, CA 91789, USA. Execution preferred in EU regions.
  • Anthropic, PBC (Claude AI model for process discovery and scoring) — 548 Market St PMB 90375, San Francisco, CA 94104, USA. The content you enter in the discovery chat is transmitted to and processed by the Anthropic API.
  • Mistral AI SAS (alternative Mistral Large AI model for process discovery and saving the process, only if your organisation — or you yourself, without an organisation — have selected this model) — 15 rue des Halles, 75001 Paris, France. The content you enter in the discovery chat, including uploaded files, is transmitted to the Mistral API and processed in data centres within the EU.
  • Gladia SAS (real-time speech-to-text transcription, only when you actively start voice input) — based in Paris, France. Audio is processed by default in EU data centres (France). See § 11 for details.
  • OpenAI Ireland Ltd. (speech synthesis — reading AI replies aloud, only when you actively switch voice mode on) — 1st Floor, The Liffey Trust Centre, 117–126 Sheriff Street Upper, Dublin 1, D01 YC43, Ireland. Only the AI's reply text is transmitted — never your audio recording. Processing takes place in the EU region with zero data retention. See § 11 for details.

Data processing agreements pursuant to Art. 28 GDPR are in place with all named providers, including their respective sub-processor lists. No further third-party disclosures occur.

8. Transfers to Third Countries (USA)

Vercel Inc. and Anthropic, PBC are based in the USA. Transfers of personal data to the USA occur in particular when using the AI chat (Anthropic).

Vercel Inc. is self-certified under the EU-US Data Privacy Framework (adequacy decision of the EU Commission dated 10 July 2023); the transfer is therefore primarily based on Art. 45 GDPR. In addition, we have concluded EU Standard Contractual Clauses (SCC) under Art. 46(2)(c) GDPR.

Anthropic, PBC is, to our knowledge, currently not self-certified under the DPF. Transfers to Anthropic are therefore based on EU Standard Contractual Clauses (SCC) under Art. 46(2)(c) GDPR, supplemented by additional technical and organizational measures (Transfer Impact Assessment).

We point out that despite these safeguards, a residual risk remains with transfers to the USA: under certain conditions, US authorities may request access to data, and EU data subjects may not have effective legal remedies under EU law against such access.

The optional speech-to-text feature (§ 11) is provided by an EU-based processor (Gladia SAS, France). No third-country transfer occurs in connection with this feature. The same applies to the Mistral Large AI model (Mistral AI SAS, France): if your organisation — or, without an organisation, you yourself — select this model, the chat content is not transmitted to Anthropic.

9. Retention Periods

  • Account and profile data: until account deletion by you or by us at contract end.
  • Chat content and process data: for as long as the account exists, but no longer than 24 months after last use, unless contractual or legal reasons require longer retention.
  • Audio and transcript data from voice input: at Gladia, per the data processing agreement (Gladia's standard retention period is up to 12 months; contractually reducible down to no retention). On our side, transcripts are stored only if you actively send them as a chat message — in that case the chat-content retention above applies.
  • Technical logs: a maximum of 30 days, then automatically deleted.
  • MCP usage data: a maximum of 180 days. Expired entries are deleted when the next new entry is written; with no activity at all this can be correspondingly later. If your account is deleted, the link to your user account as well as the client identifier and user agent are removed from your entries. Time, company, process reference, tool and outcome remain as statistics until the period ends.
  • Data with statutory retention obligation (invoices, contracts): 6 or 10 years pursuant to § 257 HGB and § 147 AO.

10. AI Processing and Automated Evaluation

The platform uses AI models by Anthropic (Claude) to run process discovery and to compute process scores.

Scoring is carried out automatically based on stored rules and AI-assisted evaluation. It does not produce legal effects on you nor similarly significantly affect you within the meaning of Art. 22(1) GDPR — the results are used exclusively for internal process evaluation at the contracting organisation. Automated decisions with legal effect or comparable impact on individuals do not take place.

Under Anthropic's Commercial Terms, content from the discovery chat is not used by default to train the generally available Claude models. Anthropic retains API requests for up to 30 days for safety and abuse prevention purposes; they are automatically deleted thereafter.

11. Speech Processing — Voice Input and Reading Aloud (optional, user-activated)

The discovery chat offers an optional voice input. You can activate it by clicking the microphone icon in the input field and disable it the same way. As long as you do not activate the feature, no audio is captured or transmitted.

Activation. On first activation, your browser will additionally ask for permission to access your microphone. Without that browser permission, no transmission occurs. You can revoke any granted permission at any time in your browser settings.

Data processed.

  • the microphone audio stream during recording (16 kHz PCM, speech only; not persisted in your browser)
  • the partial and final transcripts that Gladia generates from this audio and that are inserted into the platform's input field
  • an accompanying vocabulary of company- and process-specific terms (e.g. tool names from our tool database, system names captured in your processes) that improves recognition accuracy; this vocabulary does not contain personal data of third parties

Processor. Gladia SAS, Paris, France. A data processing agreement under Art. 28 GDPR is in place with Gladia. Audio streams are processed by default in data centres within the EU (France). No transfer to third countries occurs in connection with speech transcription.

Retention and model training. We operate the voice feature on a paid Gladia plan. Under Gladia's contractual commitments for that plan, audio and transcript data are not used to train AI models. Gladia's standard retention period is up to 12 months and can be reduced in the data processing agreement. Within the platform itself, transcripts are stored only if you actively send them as a chat message — in that case they are treated like other chat content (see §§ 4, 9).

No biometric processing. We use the audio signal exclusively to transcribe spoken content. No speaker recognition or biometric identification takes place. The processing therefore does not fall within Art. 9 GDPR (special categories of personal data).

Termination. You can end voice input at any time by clicking the microphone icon again. We then send Gladia a stop message and close the audio connection. It is at your discretion whether to subsequently send, edit or discard the transcript shown in the input field.

Legal basis. Art. 6(1)(b) GDPR: voice input is an alternative input method for the contracted discovery function. The processing is necessary to provide the input modality you selected. Because the feature only runs while you actively use it, you can prevent the processing at any time by not activating, or by deactivating, the feature.

Voice mode: replies read aloud. In addition to voice input, you can switch on a voice mode in which the AI's replies are read aloud. Only the AI's already-generated reply text is transmitted to OpenAI Ireland Ltd., Dublin, Ireland — not your audio recording, not your input, and none of your process data. OpenAI returns an audio file which your browser plays and then discards; it is not stored on our systems.

A data processing agreement pursuant to Art. 28 GDPR is in place with OpenAI Ireland Ltd. Processing takes place in the EU region with zero data retention; no third-country transfer occurs as part of speech output, and the transmitted texts are not used to train AI models. We deliberately do not use in-browser speech synthesis (e.g. Google or Microsoft voices): it would transmit the reply text to further providers not named here. If speech output is unavailable, the reply stays silent and readable only.

Legal basis and termination. Art. 6(1)(b) GDPR, as for voice input: voice mode is an output modality you select for the contracted discovery function. It is off by default, switched on only by your click, and stays off after a page reload. Clicking again switches it off at any time; no further reply text is then transmitted to OpenAI.

12. Cookies and Similar Technologies

We use only strictly necessary cookies and storage mechanisms within the meaning of § 25(2)(2) TDDDG (formerly TTDSG, renamed on 14 May 2024). No consent is required for these.

  • Supabase Auth (sb-*): session and refresh tokens, HTTP-only cookies, required to maintain login state.
  • NEXT_LOCALE: stores your language preference.
  • theme (localStorage): stores your dark/light mode selection.

Tracking, marketing or analytics cookies are currently not in use. Should this change, we will introduce an appropriate consent mechanism and inform you beforehand.

13. Right to Object (Art. 21 GDPR)

Where we process your personal data on the basis of a legitimate interest (Art. 6(1)(f) GDPR), you have the right to object to such processing at any time on grounds relating to your particular situation.

In the event of an objection, we will no longer process the data concerned unless we can demonstrate compelling legitimate grounds for the processing that override your interests, rights and freedoms, or the processing serves the establishment, exercise or defence of legal claims.

Please address your objection to: info@teclead-ventures.de

14. Further Data Subject Rights

In addition to the right to object (§ 13), you have the following rights:

  • Access (Art. 15 GDPR) to data stored about you
  • Rectification (Art. 16 GDPR) of inaccurate data
  • Erasure (Art. 17 GDPR), unless statutory retention obligations apply
  • Restriction of processing (Art. 18 GDPR)
  • Data portability (Art. 20 GDPR)
  • Complaint to a supervisory authority — the competent authority for us is the Berlin Commissioner for Data Protection and Freedom of Information

To exercise your rights, an informal message to info@teclead-ventures.de is sufficient.

15. Data Security

We implement appropriate technical and organisational measures under Art. 32 GDPR, including TLS encryption for data in transit, encrypted storage of sensitive data, access and role concepts, and regular review of our processes.

16. Changes to This Privacy Policy

We will update this privacy policy when legal requirements or the scope of the platform change. We will notify you of material changes in good time by email or within the application.