Looking for a Fellow Alternative? What to Know First
Fellow is unusual in that it combines genuine meeting management, agendas, action items and 1:1s, with a modern AI notetaker. Anyone searching from Germany for a Fellow alternative usually does so for a specific reason: the data is hosted in Canada, the AI and transcription run through US sub-processors, and the botless desktop mode is invisible to external participants. This comparison covers Fellow's real strengths, frames the data-protection question for German businesses, and sets Sally, hosted in Germany, against it.
What Fellow Does Well
Fellow has clear strengths, and its meeting-management heritage sets it apart from pure notetakers.
Meeting Management Plus Notetaking
Most competitors only take notes; Fellow also runs the meeting. Collaborative agendas, action items, meeting templates, feedback and 1:1 tooling sit alongside recording, transcription, summaries and an 'Ask Fellow' chat. For teams that want structure before and after the call, not just a transcript, that combination is genuinely useful.
Flexible Capture Across Platforms
Fellow captures via a visible note-taker bot or a botless desktop mode, and works across Zoom, Microsoft Teams, Google Meet, Webex, Slack huddles, phone and in-person meetings. That breadth means it fits a wide range of meeting setups.
A Solid Compliance Baseline
Fellow is SOC 2 Type II certified, offers a GDPR-compliant Data Processing Agreement, employs an on-staff Data Protection Officer and states it does not train AI models on customer data, with AES-256 encryption at rest. That is a respectable baseline, and better than several pure notetakers offer.
Is Fellow GDPR-Compliant? The Honest Answer
Fellow does the paperwork well, but the data path crosses two third countries, which is where it matters for German businesses.
Canada Hosting Is Better Than the US, but the AI Runs in the US
Fellow hosts its core data on AWS in Canada. Canada holds a partial EU adequacy decision, so Canadian hosting is more defensible than plain US hosting. The complication is that Fellow's transcription and AI sub-processors, Deepgram, AssemblyAI, OpenAI, Anthropic and Google among them, are US providers, and the meeting bots are run through a US service. The Canada adequacy does not cover those US sub-processors, so transcript content still reaches the United States. Whether the overall chain is permissible turns on the providers' standing under the EU adequacy framework.
No EU Residency, and a Transfer to Assess
There is no documented EU or German data-residency option, so a German organisation cannot keep the data inside the EU. Because the AI processing involves US providers, a complete Transfer Impact Assessment under Schrems II has to cover that chain, including the risk from US surveillance law. The DPA and the DPO help, but they do not remove this step.
§ 201 StGB: The Botless Mode Is Covert Toward Outsiders
Fellow's botless desktop mode captures system audio locally with no participant shown in the call. Internal Fellow users get a notification, but external guests see no indicator that recording is taking place. That is exactly the scenario § 201 StGB (violation of the confidentiality of the spoken word) targets: it prohibits the covert recording of non-publicly spoken words and carries a penalty of up to three years' imprisonment. The visible note-taker bot is the safer mode, but whichever you use, you must inform all participants and obtain consent. Sally's help center explains how consent and data protection work in practice.
Where Fellow Falls Short for German Users
Beyond the data path, two points weigh against Fellow specifically for the German market.
No German Data Residency
For regulated industries such as legal, healthcare, finance or the public sector, processing outside the EU, in Canada and the US, is frequently a disqualifying criterion no matter how good the tool is. Even where it is allowed, the SCC-and-TIA paperwork is real overhead for a small or medium-sized business. A provider that hosts in Germany removes that work entirely, a point we develop in our overview of German servers versus the US cloud.
German Language Not DACH-Tuned
Fellow handles German via automatic language detection across roughly 100 languages, but with a general multilingual model rather than DACH-specific tuning. Austrian German, Swiss German and regional dialects are recognised less accurately than standard High German, and German technical terminology remains a known challenge for models trained predominantly on English data.
Sally: The Alternative Hosted in Germany
For German businesses that want Fellow's structure without the Canada-and-US data path, Sally rests on a different footing. Sally is an AI meeting assistant from Aliru GmbH in Mannheim, hosted and developed exclusively in Germany.
Data Stays in Germany
All data is processed and stored exclusively on servers in Germany. There is no third-country transfer to assess and no Transfer Impact Assessment under Schrems II, because all processing takes place within the EU, with no Canadian hosting and no US AI sub-processor chain. The contractual partner is Aliru GmbH, a German company that signs a German data processing agreement and is liable under German law. Sally is also ISO-certified, with information security verified by independent audits rather than only declared.
German Language, Visible Bot and Native Integrations
Sally is optimised for German and German dialects rather than adapted from a general model, and it joins meetings as a visible bot so recording is transparent from the start, with no covert botless mode. For online meetings Sally joins Microsoft Teams, Zoom and Google Meet automatically; for in-person conversations the Sally App on iOS and Android records directly. Meeting notes flow into seven native CRM integrations, including HubSpot and Salesforce, and support is available in German from a German team.
Fellow vs. Sally: A Direct Comparison
An overview of the key differences at a glance:
| Criterion | Fellow | Sally |
|---|---|---|
| Product category | Meeting management + AI notetaker | Managed meeting assistant |
| Provider / contracting party | Fellow Insights Inc. (Ottawa, Canada) | Aliru GmbH (Mannheim, Germany) |
| Data storage | Canada (AWS) | Germany |
| EU data residency | No | Yes (Germany only) |
| AI / transcription sub-processors | USA (Deepgram, AssemblyAI, OpenAI, Anthropic, Google) | Processing in Germany |
| Data processing agreement (DPA) | Yes | Yes (German DPA) |
| ISO certification | None | ISO 9001, ISO 14001, ISO 27001 |
| Capture method | Visible bot + botless desktop mode | Visible bot + app |
| § 201 StGB (covert recording) | Risk in botless mode | No risk through visible bot |
| German language / dialects | General model, not DACH-tuned | Trained for German and dialects |
| Support language | English | German team |
| Pricing | Free plan; Team from 7 USD; Business from 15 USD/user/month | Starter from €8/user/month; 30 days free trial |
Conclusion: Is Fellow the Right Tool for German Businesses?
Fellow is a capable, distinctive product, and its meeting-management heritage makes it more than a notetaker. For a team that is comfortable with Canadian hosting and US AI processing, and wants agendas and action items built in, it is a strong choice with a solid compliance baseline.
For a German business, the data path is the deciding factor: core data in Canada, AI and transcription in the US, no EU residency, and a botless mode that is covert toward external participants. Add German-language quality that is not DACH-tuned and English-only support, and a German-hosted solution becomes the cleaner fit.
Anyone who wants comparable capability with the data kept in Germany will find Sally a GDPR-compliant alternative, developed and hosted in Germany, with German support, native integrations and a visible bot that keeps recording transparent. Sally can be tested free of charge for 30 days; see the pricing page for details.
Disclaimer: This is not legal advice.
