Ask Fellow: Meetings aren't just "recorded" anymore—they've started "working"

2026-02-25 | ProductHunt | Official Website

Fellow's meeting interface: The AI note-taker appears as a "fourth participant" in the video grid, with the agenda, action items, and discussion points synced on the right. This design makes the AI feel like a team member rather than a background tool.

30-Second Quick Judgment

What it does: The core AI Agent functionality of Fellow.ai—it doesn't just take notes; it writes follow-up emails, updates your CRM, generates documentation, and clips video highlights. Essentially, it automates all the "annoying post-meeting chores."

Is it worth watching?: Absolutely. In the crowded AI meeting assistant market, most products stop at "taking notes." Ask Fellow jumps straight to "doing the post-meeting work." This is an evolution from tool to Agent, and it's the right direction. Note: The free version only offers 5 uses, pricing is premium, and it requires team-wide adoption to unlock its full value.

Three Questions for You

Is it relevant to me?

Who is the target user?:

• Middle managers with 3+ meetings daily (1:1s, team syncs, project updates).
• Sales teams (needing follow-up emails and CRM updates).
• Customer Success teams (tracking client commitments).
• Remote/hybrid teams.

Is it me?: If you spend over 2 hours a week in the loop of "organizing notes → sending follow-ups → updating task managers," you are the target user. If you're a solo dev or freelancer with few meetings, this product might be overkill.

When would I use it?:

• After a Monday sync → Ask Fellow automatically drafts the summary for the whole team.
• After a client call → One-click push of key info to Salesforce.
• During performance review season → Ask it to draft an evaluation based on the last six months of 1:1 records.
• When NOT to use it: Casual 2-person brainstorms or off-the-record chats.

Is it useful to me?

ROI Judgment: If you manage a team of 10+ and have 10+ meetings a week, the investment pays off. For a small 3-person team, the free versions of Fathom or tl;dv might suffice.

Is it delightful?

The "Aha!" moments:

• Cross-meeting Search: Ask "What goals did Sarah set last quarter?" and it pulls from dozens of records. This is far superior to manual searching.
• One-click Follow-ups: Just say "write a follow-up email," and it packages the summary and next steps perfectly.
• Video Snippets: Want to share a key moment? Have it clip the segment instead of searching for timestamps yourself.

Real User Feedback:

"The meeting to content to action loop is underrated. Love how Ask Fellow connects those dots." — @Ivamotion

"The briefing feature sometimes misidentifies content, sending irrelevant information." — Long-term user review from thebusinessdive.com

For Independent Developers

Tech Stack

• Frontend: Web app (Official site built with Framer)
• Backend: AWS (Canada Central)
• AI/Model: Anthropic Claude (Explicitly states user data is not used for training)
• Encryption: AES 256-bit at rest, HTTPS in transit
• Integrations: 50+ apps (Zoom, Google Meet, MS Teams, Slack, Jira, Asana)

Core Implementation

Ask Fellow's core is a RAG (Retrieval-Augmented Generation) system: it indexes all meeting transcripts. When a user asks a question, it retrieves relevant snippets and uses an LLM to generate an answer. Email generation, doc creation, and CRM updates are downstream tasks built on this foundation. The 500+ templates and collaborative agendas represent traditional SaaS product depth.

Transcription accuracy is a key competitive edge: Fellow reaches 95%+, significantly higher than Otter.ai's ~85%, especially in noisy environments or with accents.

Open Source Status

• Is it open source?: No. Fellow has 60 repositories on GitHub (github.com/fellowapp), but they are mostly integration tools and helper libraries.
• Similar Open Source Projects: Meetily (Privacy-first AI meeting assistant using Whisper).
• Build Difficulty: High. Basic transcription + summary is easy (Whisper + GPT), but the moat is the cross-meeting search index, 50+ integrations, and enterprise compliance (SOC2/HIPAA/GDPR), which requires a 3-5 person team and 12+ months.

Business Model

• Monetization: SaaS Subscription
• Pricing: Free (5 uses) → Solo ($19/mo) → Pro → Enterprise
• Annual Revenue: ~$15M (as of August 2025)
• Team: ~61 people across 3 continents

Big Tech Risk

High risk. Microsoft Teams has Copilot, Google Meet has Gemini, and Zoom has AI Companion. The giants are all moving into this space. Fellow's moat is: (1) Cross-platform support; (2) Stricter security/compliance; (3) Legacy strength in meeting management (agendas/templates). Long-term, if giants perfect their Agent capabilities, Fellow's space may shrink.

For Product Managers

Pain Point Analysis

• Problem Solved: The disconnect in post-meeting execution—things are said, but never followed up on.
• Severity: High-frequency, high-pain. Knowledge workers spend roughly 31 hours a week on meetings and follow-ups. "Recording" only solves 30% of the problem; the real pain is turning talk into action.

User Persona

• Core User: Middle managers with 5-15 direct reports and 15+ weekly meetings.
• Secondary User: Sales AEs and CSMs needing precise client follow-ups.
• Scenarios: Recurring 1:1s, project checkpoints, client calls, town halls.

Feature Breakdown

Competitive Differentiation

Key Takeaways

1. Evolution from "Record" to "Action": Don't stop at data collection; help users turn data into outcomes.
2. Security as a Differentiator: In the age of AI anxiety, "we don't train on your data" is a powerful selling point.
3. Collaborative Agendas: Create value before the meeting starts, not just after.

For Tech Bloggers

Founder Story

Aydin Mirzaee is a serial entrepreneur. Along with his brother Amin Mirzaee and CTO Samuel Cormier-Iijima, he previously built Fluidware (online surveys), which was acquired by SurveyMonkey. After a stint there, the trio founded Fellow in 2017.

Interestingly, Fellow underwent a deep transformation: in 2019, it was just an agenda and notes tool. Over the last two years, this Ottawa-based company pivoted everything—product, pricing, and market strategy—to become an AI-first company. In the CEO's words, it's a metamorphosis from a "meeting tool" to an "AI Meeting Agent."

Discussion Angles

• The "Intrusiveness" of Meeting Bots: Many find it uncomfortable to have a bot recording, especially in sensitive 1:1 sales scenarios.
• The "Trial Trap": 5 uses aren't enough to evaluate the product, especially compared to more generous competitors.
• The Network Effect Dilemma: It requires company-wide adoption to be truly valuable, which is a high barrier to entry.

Buzz Data

• PH Ranking: 282 upvotes (Moderate buzz)
• Twitter Discussion: Low interaction on launch day (peak 4 likes)
• G2 Reviews: Consistently positive, especially regarding security and accuracy.

Content Suggestions

• Angle: "AI Meeting Assistants are evolving from 'recorders' to 'agents'—what does this mean for the future of work?"
• Trend Opportunity: AI Agents are the hot topic; Fellow is a rare example of an Agent actually landing in an enterprise setting.

For Early Adopters

Pricing Analysis

Getting Started

• Setup Time: 30 mins for basic config, 1-2 hours to master features.
• Learning Curve: Medium-high (lots of features and concepts).
• Steps:
  1. Sign up at fellow.ai, connect Google Calendar or Outlook.
  2. Choose which meetings to auto-join (start with internal ones).
  3. After your first meeting, test the summary and email generation.
  4. Try cross-meeting search with a question that spans multiple sessions.

Pitfalls & Complaints

1. Stingy Free Tier: 5 uses and you're forced to pay, unlike Otter's 300 mins/month.
2. Bot Awkwardness: External clients will see "Fellow AI" join; you may need to explain it beforehand.
3. Action Item Accuracy: AI-identified tasks aren't always perfect; manual review is still needed.
4. Requires Scale: If you're the only one using it, the collaborative features lose their value.
5. Resource Heavy: The desktop app can be laggy on older machines.

Security & Privacy

• Storage: AWS Canada Central.
• Privacy: No training on user data; data is deleted after processing.
• Audits: SOC 2 Type II + HIPAA + GDPR.
• Admin Controls: Granular settings for who can record, view replays, and data retention.
• Highlight: Auto-delete policies (as short as 1 day) and the ability to pause/resume recording.

Alternatives

For Investors

Market Analysis

• Market Size: AI Meeting Assistant market was $2.4B in 2024, projected to reach $15.2B by 2032.
• Growth Rate: 25-35% CAGR.
• Drivers: Normalization of hybrid work, rapid AI advancement, and enterprise anxiety over meeting efficiency.

Competitive Landscape

Timing Analysis

• Why Now: The AI Agent concept is moving from hype to reality; "post-meeting automation" is finally technically viable.
• Tech Maturity: Larger LLM context windows make cross-meeting search possible; Function Calling makes CRM updates reliable.
• Market Readiness: After 3 years of remote work meeting fatigue, the willingness to pay for "meeting tax reduction" is at an all-time high.

Team Background

• Founders: Aydin Mirzaee (Serial entrepreneur, Fluidware → SurveyMonkey → Fellow).
• Core Team: ~61 people across North America, Europe, and South America.
• CTO: Samuel Cormier-Iijima (Long-time partner of the CEO).

Funding Status

• Total Raised: $30.5M (Seed + Series A).
• Investors: Inovia Capital (Series A lead) + 10 institutional investors.
• Revenue: ~$15M ARR (2025 estimate).
• Valuation: Undisclosed.

Conclusion

Ask Fellow represents the evolution of the AI meeting space: from "helping you remember" to "helping you do." While the direction is correct, execution faces pressure from tech giants, a stingy free tier, and the hurdle of team-wide adoption.

Resource Links

2026-02-26 | Trend-Tracker v7.3