Review

Fellow: meeting memory for teams that care about control

Fellow is a strong choice for teams that need meeting notes, follow-up workflows, and governance in one place, but its best value only appears once the whole workspace uses it.

Last updated April 2026 · Pricing and features verified against official documentation

Fellow is what happens when a meeting assistant stops pretending that transcription is the whole job. The product now sits closer to meeting infrastructure than to a simple notetaker: it records calls, generates summaries and action items, supports bot and botless capture, stores notes in a shared workspace, and routes the output into the rest of a company’s stack.

That matters because the market has split. One camp wants a lightweight recorder that quietly spits out a summary. The other wants a governed system for keeping meeting context, follow-up, and access control in one place. Fellow belongs to the second camp, and it is strongest when the organization already treats meetings as a workflow, not just a calendar event.

The honest case for Fellow is straightforward. It is a good buy for teams that need structured meeting output, shared visibility, and admin controls without sacrificing cross-platform coverage. It is especially compelling for operations, sales, customer success, HR, and compliance-heavy teams that want the notes to become part of the work.

The honest case against it is just as clear. Fellow is not the calmest or cheapest meeting tool on the shelf, and the product only earns its keep when enough people actually adopt it. If you want a personal note-taker more than a team system, Fellow can feel like more machine than you need.

What the Product Actually Is Now

Fellow is no longer just a recorder with summaries attached. It is a meeting workflow platform that spans pre-meeting briefs, agendas, note-taking, recordings, action items, Ask Fellow chat, CRM updates, and workspace-level controls. The product is built to make meetings searchable and reusable across Zoom, Google Meet, Microsoft Teams, Slack huddles, and in-person conversations.

That also explains the product’s personality. Fellow is built around the idea that meeting data should be owned, permissioned, and operationalized inside a shared workspace. The result is more admin-friendly than consumer note apps and more opinionated than a generic AI chatbot.

Strengths

It turns meetings into shared operating memory. Fellow’s core strength is not just producing a transcript, but making that transcript useful afterward. Summaries, action items, searchable history, and Ask Fellow all help teams recover decisions and context without digging through raw recordings.

It handles governance better than most meeting tools. Fellow’s current posture is unusually explicit: it says its AI is never trained on customer data, and the product exposes controls for access, retention, recording behavior, and voice matching. That is the kind of setup security-minded teams actually need when they are deciding whether to allow a meeting assistant at all.

It works across the platforms real teams use. Zoom, Google Meet, Microsoft Teams, and Slack huddles are all covered, and the product supports both visible bot capture and botless audio capture. That flexibility matters because it avoids the usual problem where a note-taker is great in one meeting stack and awkward everywhere else.

Its follow-up story is stronger than its summary story. Fellow is more useful when the output needs to flow into Salesforce, HubSpot, Notion, Confluence, Jira, Asana, or internal workflow systems. Meeting notes are easy to generate; making them actionable is the harder problem, and Fellow is trying to solve that problem instead of just the note itself.

Weaknesses

The free tier is more of a sample than a plan. The current pricing page caps Free at five seats and only five AI notes and five AI recordings per user for life. That is enough to evaluate the product, but it is not enough to build a habit around it.

The product gets more valuable only after rollout. Fellow’s strongest features depend on team adoption, shared permissions, and downstream workflow connections. If only one person uses it, you mostly have a personal meeting archive; the real payoff comes when the workspace standardizes on it.

It is narrower than a general-purpose AI workspace. Fellow can draft follow-ups and answer questions from meetings, but it is still a meeting-centric product. If your main need is broad research, writing, or coding help, you should look elsewhere rather than paying for a specialist system that happens to include chat.

Pricing

Fellow’s pricing makes sense once you stop reading it like a consumer app and start reading it like workspace software. Free is useful for evaluation, Team is the first tier that feels genuinely practical for individuals or small groups, Business is the value tier for most teams, and Enterprise is where procurement, security, and admin needs take over.

The current official pricing page shows Free at $0, Team at $7 per user per month billed annually or $11 monthly, Business at $15 per user per month billed annually or $23 monthly, and Enterprise at $25 per user per month billed annually with sales contact for larger deployments. The main trap is the same one that catches a lot of meeting software: the annual pricing is the real price, while monthly billing is meaningfully more expensive.

For most serious users, Team is the sensible individual or small-group choice because it unlocks the product’s real workflow value without forcing an enterprise buy. Business is the better value for departments because the unlimited notes and recordings, sales templates, and broader admin posture make it feel like a company system rather than a personal subscription.

Privacy

Fellow’s privacy stance is better than average for the category, and it is one of the main reasons the product exists. The company says its AI is never trained on customer data, the help center says user-entered notes and integration data are used only to provide the service, and the current privacy policy frames Fellow as a data processor for organizational customers rather than a general-purpose data broker.

The tradeoff is that Fellow still collects and processes a lot of sensitive meeting infrastructure: transcripts, notes, calendar data, usage analytics, and, if voice matching is enabled, voice samples for speaker labeling within a workspace. Voice matching is opt-in and can be disabled, but it is still biometric processing, so teams in regulated environments should treat that feature as a real policy decision rather than a convenience toggle.

On compliance, Fellow says it is SOC 2 Type II, HIPAA, and GDPR compliant, with enterprise-grade encryption and access controls. That puts it in the right conversation for healthcare, finance, and legal teams, but it does not remove the need to check internal retention rules, recording policies, and who can access what inside a workspace.

Who It’s Best For

• Operations teams that need meeting discipline. Fellow is a fit for chief-of-staff, ops, and program-management roles that want agendas, notes, action items, and follow-up routed through one shared system instead of scattered across documents and chat.
• Sales and customer-success teams with CRM hygiene problems. If call recap quality matters and meeting notes need to reach Salesforce or HubSpot without manual copying, Fellow is stronger than a generic recorder because it treats downstream workflow as part of the product.
• Security-conscious companies rolling out AI meeting tools company-wide. Fellow makes sense when IT or compliance wants explicit access controls, no-training guarantees, and workspace-level governance before allowing an assistant into meetings.
• Hybrid teams that meet on multiple platforms. If some calls happen in Zoom, some in Meet, some in Teams, and some in Slack huddles, Fellow’s cross-platform capture is easier to standardize than juggling multiple native assistants.

Who Should Look Elsewhere

• People who want a lighter personal note-taker should compare Fathom first. It is a cleaner fit if you mostly want transcripts and summaries without rolling out a shared system.
• Teams that live on high-volume follow-up automation should compare Fireflies.ai. Fireflies is more aggressively platform-like and may fit better if your primary goal is routing meeting output into many downstream automations.
• Users who want a broad AI workbench should start with ChatGPT instead. Fellow is too meeting-specific to replace a general assistant for writing, research, or coding.
• Teams that want a narrower collaboration layer should look at tldv. It is easier to justify if your main job is recording and recap, not workspace governance.

Bottom Line

Fellow is one of the better examples of an AI meeting tool that understands the actual problem. The problem is not producing text; it is creating a governed record of what happened, who owns the next step, and where that context should live after the call ends. On that axis, Fellow is credible.

It is also honest about the cost of that ambition. Fellow is most valuable when a team commits to it, and that makes it harder to recommend as a casual convenience. If meetings are part of your operating system, Fellow is a serious candidate. If not, it will feel like infrastructure you do not need yet.