“Leading partner for AI assistants for Swiss network operators and fiduciary firms”
Connect AI develops and operates tailor-made AI assistants for companies, primarily for customer service, internal knowledge work, sales, and industry-specific scenarios such as fiduciary services. The core product, Employee GPT, is marketed as a secure alternative to ChatGPT/Copilot for companies that can be operated in Switzerland or the EU. The offering is complemented by AI rollout programs for executives, teams, and companies.
Connect AI
AI assistants for businesses, especially for customer service, internal knowledge work, sales, and industry-specific scenarios such as fiduciary services
Location: Switzerland ⓘ Connect AI Group GmbH, Lagerstrasse 93, 8004 Zurich, Switzerland
AI Customer Support AI agents, help center, personalized responses, knowledge library, analytics and reports, service quality, and multilingual 24/7 support.
Sales Agent Personal sales agent for customers, lead qualification, product recommendations, upselling/cross-selling, audience insights, analytics, and reduction of cart abandonment.
Event Agent / TreuhandGPT / AI Strategy Listed on the website as additional solutions; specific public details on plans and features are not fully verifiably available. Other Customer-specific implementation & operation The Terms and Conditions describe Connect-AI Services as the customer-specific implementation and operation of the Connect-AI platform; fees, packages, usage volume, and term are defined in the respective Agreement.
Onboarding / Consulting / Integrations According to the Employee-GPT blog, the pricing framework includes not only the platform and secure hosting but also training and personal support; specific terms are clarified on a project-by-project basis.
“Your AI partner for solutions with real business impact”
Video currently unavailable
Target audience
Connect AI is aimed primarily at companies in the DACH region, especially SMEs and larger organizations that want to do more than just experiment with AI and instead integrate it into processes in a controlled way. The website explicitly addresses industries such as Swiss network operators and fiduciary firms, but also showcases use cases for internal knowledge work, customer service, sales, and events. Typical user roles include executive management, department heads, customer service teams, sales managers, as well as IT and AI managers.
Outstanding features
Connect AI is particularly strong where a standard AI is not enough. Employee GPT combines internal knowledge use, multi-level access rights, company-specific assistants, and multi-LLM support. In addition, there are specialized solutions such as AI Customer Support with knowledge library and analytics, Sales Agent with personalization and upselling, as well as TreuhandGPT with focused specialist agents for tax- and fiduciary-related tasks. The fact that Connect AI also provides a publicly accessible DPA including a subprocessor list is a plus point compared to the market for procurement and compliance reviews.
Main areas of application
The strongest areas of application are internal company chat with proprietary knowledge, secure support for employees, 24/7 customer service, AI-supported sales assistance, and industry-specific assistance systems. Connect AI is also suitable for companies that want to accompany AI rollout, training, and governance organizationally instead of merely purchasing a tool. The platform is particularly convincing where data protection, brand/process customization, and structured implementation play a bigger role than maximum self-service simplicity.
Usage & notes
Connect AI is not a typical plug-and-play consumer tool. The offering is clearly B2B-focused and combines software with consulting, setup, integration, and adoption. For data protection reviews, one should not only read the marketing claims but definitely also consult the DPA: storage locations, deletion periods, subprocessors, and possible third-country transfers are described much more concretely there. It is also important that the general privacy policy still contains older wording and even points out unencrypted transmission for website/form data. For a reliable assessment, the specific product setup should therefore always be reviewed.
| Target audience | Assessment |
|---|---|
| SMEs in Switzerland / DACH | Very suitable – Connect AI is clearly aimed at companies that want to use AI assistants for customer service, sales, events, or internal employee support. |
| Customer service teams | Very suitable – the “AI Customer Support” solution automates support, help center, AI agents, knowledge library, analytics, and reports. |
| Sales and e-commerce teams | Suitable to very suitable – the Sales Agent supports purchasing processes, lead qualification, product recommendations, upselling/cross-selling, and reducing cart abandonment. |
| Companies with sensitive data | Very suitable – especially for fiduciary, finance, legal, healthcare, and other industries where Swiss data hosting, confidentiality, and data sovereignty are important. |
| Internal knowledge work / Employee GPT | Very suitable – Employee GPT is positioned as a secure private ChatGPT alternative for companies that want to use internal documents, processes, and knowledge databases. |
| Private individuals | Not suitable – Connect AI is a B2B provider of customized AI solutions, not a general consumer AI tool. |
Hosting & Data
1) On-prem / local hosting
Meaning: The company operates the solution on its own hardware or within its own infrastructure. In the strictest sense, not only the application runs locally, but ideally the model as well.
2) Private cloud / data center
Meaning: The solution runs in a dedicated or more clearly separated cloud environment, often with a hosting provider or hyperscaler, but in a German data center or in a particularly controlled environment.
3) EU SaaS / managed
Meaning: The provider operates the solution itself as a service. The company uses the tool as a ready-made cloud service, ideally with EU data residency.
4) Hybrid
Meaning: One part of the processing remains internal / local / in a private cloud, while another part runs in an external cloud or EU SaaS.
5) AVV / DPA
Meaning: This is the data processing agreement or Data Processing Addendum. It governs that the provider processes personal data on behalf of the customer and is bound by the customer's instructions.
6) No training
Meaning: The provider does not use your prompts, uploads, attachments, chat histories, or outputs for training or improving the general model — ideally excluded by contract.
7) Open-source / transparency path
Meaning: There is a path toward greater technical transparency and sovereignty, for example through:
- open models
- documented components
- self-hostable parts
- traceable architecture
- export / switching options
| On-prem / local hosting | ⚠️ |
| Private cloud / data center | ⚠️ |
| EU SaaS / Managed | ⚠️ |
| Hybrid | ❓ |
| DPA / AVV | ✅ |
| No training on customer data | ⚠️ |
| Open source / transparency path | ⚠️ |
On-prem / local hosting: partially
On the Employee-GPT page, the pricing section lists “Hosting Preference (Cloud or On-Premise).” This suggests an on-premises/local option. However, the website does not describe in detail whether the solution, strictly speaking, runs entirely on the customer’s own hardware, including a local model.
Private Cloud / Data Center: Partially
The website mentions “role-based access rights and private container environments” as well as “hosting with FINMA-certified partners.” This suggests isolated, controlled environments. However, the website does not clearly state whether this is offered as a dedicated private cloud architecture for customers by default.
EU SaaS / Managed: Partially
The website promotes a managed service with data storage “in Switzerland or the EU” and mentions Swiss/EU hosting for certain models. At the same time, specific details are missing regarding EU data residency by operating model, all subprocessors, and guaranteed EU/EEA locations. Therefore, this is only partially covered.
Hybrid: Indirect / Not Available
The combination of self-hosted open-source models, API access to closed models, and “cloud or on-premises” suggests possible hybrid setups. However, the website does not specifically describe an explicitly defined hybrid architecture with a division of processing between internal and external sources.
DPA: Covered
There is a separate “Data Processing Agreement” subpage on the website with a downloadable DPA PDF. This confirms the presence of a DPA on the provider’s website.
No Training: Partially
The Employee-GPT page explicitly states “No use of customer data for model training” and “EmployeeGPT does not store or further use your input/output data.” At the same time, the DPA page states that “User-provided information may be stored and analyzed for further development and customized offers.” This does not appear entirely consistent; a robust, contractually defined opt-out/no-training mechanism applicable across Europe is not fully documented on the website.
Open Source / Transparency Path: Partial
It mentions “self-hosted open-source models,” including Mistral, Gemma, and Deep Seek, which opens up a path to transparency and sovereignty. The website does not provide further details on which specific components are open-source, which parts are self-hostable, or how extensive the export and migration options are.
Data Processing
Data processing is presented in two parts on the website: From a product perspective, Connect AI describes Employee GPT as a privacy-focused offering with data stored in Switzerland or the EU, Swiss infrastructure, optional self-hosted open-source models, private container environments, and a hosting preference of “cloud or on-premises.” The general privacy policy, on the other hand, describes broader corporate and web tracking/communication processing, with physical data stored in Switzerland and Europe, electronic customer data stored on Google Cloud, and CRM usage via HubSpot, which, according to the website, stores data in the U.S., among other locations; it also mentions possible data transfers outside Switzerland/the EU/the EEA.
Conclusion
Based on the website, Connect AI cannot be generally classified as fully GDPR-compliant for users throughout the EU/EEA; however, there is a plausible path to data-protection-friendly use: controlled hosting in Switzerland or the EU, possibly even “on-premises,” plus self-hosted open-source models. Because key evidence for a robust overall EU/EEA assessment is missing or inconsistent—particularly regarding transfers to third countries, subprocessors, exact data center locations, and certifications—the overall assessment is “conditional.”
Sources
| On-prem / local hosting | ⚠️ |
| Private cloud / data center | ⚠️ |
| EU SaaS / Managed | ⚠️ |
| Hybrid | ❓ |
| DPA / AVV | ✅ |
| No training on customer data | ⚠️ |
| Open source / transparency path | ⚠️ |
On-prem / local hosting: partially
On the Employee-GPT page, the pricing section lists “Hosting Preference (Cloud or On-Premise).” This suggests an on-premises/local option. However, the website does not describe in detail whether the solution, strictly speaking, runs entirely on the customer’s own hardware, including a local model.
Private Cloud / Data Center: Partially
The website mentions “role-based access rights and private container environments” as well as “hosting with FINMA-certified partners.” This suggests isolated, controlled environments. However, the website does not clearly state whether this is offered as a dedicated private cloud architecture for customers by default.
EU SaaS / Managed: Partially
The website promotes a managed service with data storage “in Switzerland or the EU” and mentions Swiss/EU hosting for certain models. At the same time, specific details are missing regarding EU data residency by operating model, all subprocessors, and guaranteed EU/EEA locations. Therefore, this is only partially covered.
Hybrid: Indirect / Not Available
The combination of self-hosted open-source models, API access to closed models, and “cloud or on-premises” suggests possible hybrid setups. However, the website does not specifically describe an explicitly defined hybrid architecture with a division of processing between internal and external sources.
DPA: Covered
There is a separate “Data Processing Agreement” subpage on the website with a downloadable DPA PDF. This confirms the presence of a DPA on the provider’s website.
No Training: Partially
The Employee-GPT page explicitly states “No use of customer data for model training” and “EmployeeGPT does not store or further use your input/output data.” At the same time, the DPA page states that “User-provided information may be stored and analyzed for further development and customized offers.” This does not appear entirely consistent; a robust, contractually defined opt-out/no-training mechanism applicable across Europe is not fully documented on the website.
Open Source / Transparency Path: Partial
It mentions “self-hosted open-source models,” including Mistral, Gemma, and Deep Seek, which opens up a path to transparency and sovereignty. The website does not provide further details on which specific components are open-source, which parts are self-hostable, or how extensive the export and migration options are.
Data Processing
Data processing is presented in two parts on the website: From a product perspective, Connect AI describes Employee GPT as a privacy-focused offering with data stored in Switzerland or the EU, Swiss infrastructure, optional self-hosted open-source models, private container environments, and a hosting preference of “cloud or on-premises.” The general privacy policy, on the other hand, describes broader corporate and web tracking/communication processing, with physical data stored in Switzerland and Europe, electronic customer data stored on Google Cloud, and CRM usage via HubSpot, which, according to the website, stores data in the U.S., among other locations; it also mentions possible data transfers outside Switzerland/the EU/the EEA.
Conclusion
Based on the website, Connect AI cannot be generally classified as fully GDPR-compliant for users throughout the EU/EEA; however, there is a plausible path to data-protection-friendly use: controlled hosting in Switzerland or the EU, possibly even “on-premises,” plus self-hosted open-source models. Because key evidence for a robust overall EU/EEA assessment is missing or inconsistent—particularly regarding transfers to third countries, subprocessors, exact data center locations, and certifications—the overall assessment is “conditional.”
Sources
Strengths & weaknesses at a glance
| Strengths | Weaknesses |
|---|---|
| • Clear focus on companies with data protection and governance requirements. | • No standardized, transparent platform pricing tier publicly visible for all solutions. |
| • Tailored AI agents instead of just a generic chatbot. | • Strong consulting/project character; less low-threshold than typical self-service SaaS tools. |
| • Multiple practical solutions: internal AI, customer service, sales, events, fiduciary services. | • The data protection situation depends on the configuration; according to the DPA, US/UK subprocessors may also be involved depending on the setup. |
| • Publicly accessible DPA with audit, deletion, and subprocessor provisions. | • The general privacy policy is older (as of January 2024) and still mentions Privacy Shield. |
| • Multi-LLM strategy including open-source and proprietary models. | • The privacy policy explicitly states that data may be transmitted via email/web form without encryption. |
| • Adoption/enablement programs for leadership, teams, and enterprise. |
Reviews
0 reviews in total
There are no confirmed reviews for this tool yet.
Submit review
Your review will only become visible after email confirmation. This protects the portal against abuse.
Report review
Please select the reason why this review should be checked.
GDPR-compliant usage possible?
The website contains several positive statements regarding the privacy-friendly use of Employee GPT in the European context, particularly regarding data storage in Switzerland or the EU, self-hosted open-source models, “cloud or on-premises,” and “no use of customer data for model training.” However, for an assessment covering the entire EU/EEA area, the documentation is not sufficient to warrant an unqualified “yes,” because the general privacy policy also mentions data processing outside the EU/EEA, and there is no clear, detailed evidence—freely accessible on the website—regarding EU/EEA data residency, a list of subprocessors, or certifications. The best available approach appears to be a controlled deployment close to the EU/Switzerland via Swiss/EU hosting or “on-premises,” but this is only partially documented.
Positive
Positive aspects include the product-related statements regarding Employee GPT: “Your data is anonymized and remains in Switzerland or the EU,” “No use of customer data for model training,” “role-based access rights and private container environments,” “Hosting with FINMA-certified partners,” as well as the statement that self-hosted open-source models and hosting preferences—“cloud or on-premises”—are supported.
Negative
A negative or limiting aspect is that the general privacy policy also mentions transfers to third countries and lists Google Cloud and HubSpot; furthermore, it states that recipients may be located outside Switzerland/the EU/the EEA. A publicly accessible, specific list of subprocessors, reliable details on EU data center locations, a clearly described EU data residency model for all variants, and relevant certifications such as ISO 27001 or SOC 2 are not provided on the website.
Server Location
Employee GPT product page: Data remains “in Switzerland or the EU”; it also mentions “Swiss infrastructure” and, for the Swiss models, “no data transfer abroad.” General Privacy Policy: physical data is stored in “Switzerland & Europe,” while electronic data from customers and users is stored on Google Cloud; HubSpot stores data in the U.S., among other locations. Specific individual EU/EEA data center countries are not listed on the website.