Skip to content
YC Root AccessYC Root Access

Recall.ai: Unlocking the World’s Conversations

Recall.ai is building the fundamental data layer for AI — the API that powers meeting and conversation recording across Zoom, Teams, Meet, phone calls, and more. Fresh off raising $38M in Series B funding at a $250M valuation, Recall.ai is cementing its role as the infrastructure every AI company needs. With less than 30 employees, they’re running infrastructure at massive scale: powering over 1,000 companies, handling three terabytes per second of video, and reaching nearly $20M in ARR. In this interview, co-founder David shares the journey from a 19-year-old YC hackathon winner who skipped his college exams for an interview, to grinding through 120 investor meetings, to building the infrastructure that now powers over a thousand AI companies. Learn more about Recall.ai: https://www.recall.ai Chapters: 01:00 – What Recall.ai Does Today 02:20 – Running Infrastructure at Massive Scale 04:00 – From 19-Year-Old Hackathon Winner to YC 06:00 – Early Co-Founder Changes & Finding Amanda 08:00 – Building the First Call Recorder 10:00 – Pivoting to an API for Conversation Data 12:00 – 120 Investor Meetings for Seed 14:30 – Conviction From Living the Problem 17:00 – Landing First Customers & Growth to 1,000+ Companies 19:30 – Sales Lessons: $2M Pure Outbound 22:00 – $20M ARR With Less Than 30 People 24:00 – Why Conversation Data Is the Future of AI 26:00 – What’s Next for Recall.ai 28:00 – Advice to Founders: Don’t Give Up

Davidguest
Sep 9, 202526mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Recall.ai powers conversation recording infrastructure for 1,000+ AI companies

  1. Recall.ai provides an API to capture real-time audio/video from meetings, calls, and in-person conversations, serving 1,000+ companies with nearly $20M revenue and <30 employees.
  2. The company originated as a call-recording product, where reliability demands forced deep investment in infrastructure that later became Recall’s core defensible asset.
  3. A pivotal shift occurred as LLM capabilities accelerated, revealing that many AI products needed standardized access to conversation data more than another end-user recorder.
  4. Fundraising was unusually grindy—$160K after four months for the first product and a seed requiring 120 investor meetings—highlighting perseverance and founder-led networking.
  5. Growth came from founder-led sales and a talent strategy focused on rare “full-stack” high-agency hires who can own systems, product decisions, and customer outcomes.

IDEAS WORTH REMEMBERING

5 ideas

Infrastructure reliability becomes the moat when data is unrecoverable.

Recording failures permanently lose customer conversations, so Recall spent 70–80% of engineering time on stability and scale—work that later compounded into a differentiated platform others didn’t want to rebuild.

Pivot timing followed platform shifts in LLM capability.

As LLMs made unstructured conversation data newly valuable, Recall repositioned from an end-user recorder to an API layer that many AI and SaaS products needed to ship faster.

Deep conviction is strongest when you’ve personally carried the pager.

David’s “nightmares” from being on-call and handling angry customers created visceral certainty the problem was real, helping sustain the team through skepticism and long fundraising cycles.

Your first customers may be former competitors if you remove their pain.

Recall emailed rival call-recording companies offering to take the infrastructure burden; early adopters gained speed advantages, which then pressured others to follow.

Sales is learnable and mostly operational—not ‘black magic.’

David frames sales as (1) clearly explaining value and (2) guiding buyers through internal procurement; Recall’s first $2M revenue was pure outbound led by the founders.

WORDS WORTH SAVING

5 quotes

We are today nearly at $20 million in revenue with a team of less than 30 people.

— David

We're launching 8 million EC2 instances every month.

— David

For the five or six-month period after we launched that call recording product... I actually had nightmares about the infrastructure.

— David

The first $2 million in revenue... was pure outbound and 100% done by Amanda.

— David

There's five times more words spoken at work every year than all the words on the entire internet.

— David

Meeting recording API for developersMassive-scale AWS infrastructure (EC2, video throughput)Founder journey: hackathon win to YC acceptanceCo-founder transitions and commitment alignmentPivot from call recorder to conversation-data infrastructureSeed fundraising via cold founder outreachFounder-led outbound enterprise salesLean team and high-agency hiring philosophyConversation data as critical AI contextProduct expansion: botless desktop SDK, phone/mobile recording, storage/query/preprocessing

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome