Jason Unpacks Sequoia’s New Funds | E2199

October 28, 2025

81 min episode · 2 min read

Episode

81 min

Read time

2 min

Topics

Fundraising & VC

AI-Generated Summary

Published Dec 25, 2025

Key Takeaways

✓Fund Size Discipline: Sequoia raised $750M for Series A (modest increase from historical $350-500M) and $200M for seed, maintaining discipline while competitors expand aggressively. Partners limited to three deals annually, ensuring concentrated positions and higher expected returns through selective investment strategy.
✓AI Infrastructure Spending: Meta, Alphabet, Microsoft, and Amazon earnings reports reveal critical data about AI infrastructure investment continuation into 2026. Founders should monitor these hyperscaler announcements for industry buying patterns and compute capacity constraints that directly impact AI startup opportunities and funding availability.
✓Self-Driving Data Strategy: Tesla uses fully synthetic video generation to train FSD systems, creating unlimited edge case scenarios like deer crossings without real-world capture. Wave and Waymo employ similar approaches, suggesting synthetic data generation becomes standard methodology for autonomous vehicle development and safety validation.
✓Expense Fraud Prevention: AI-generated fake receipts create new fraud vectors, but companies like Ramp deploy AI detection systems in response. Founders should implement stipend-based expense systems ($50-75 daily) rather than receipt tracking to eliminate fraud incentives and reduce administrative overhead while maintaining employee trust.
✓Marketplace Revenue Reporting: Companies like Mercor paying 60-70% of gross revenue to service providers face scrutiny over ARR calculations. Founders must distinguish between gross marketplace volume and net retained revenue when reporting metrics to investors, as conflating these figures creates misleading valuation multiples and funding expectations.

What It Covers

Jason Calacanis analyzes Sequoia Capital's new $750M Series A and $200M seed funds, emphasizing their disciplined approach with limited partner involvement and concentrated portfolio strategy versus competitors raising larger funds.

Key Questions Answered

•Fund Size Discipline: Sequoia raised $750M for Series A (modest increase from historical $350-500M) and $200M for seed, maintaining discipline while competitors expand aggressively. Partners limited to three deals annually, ensuring concentrated positions and higher expected returns through selective investment strategy.
•AI Infrastructure Spending: Meta, Alphabet, Microsoft, and Amazon earnings reports reveal critical data about AI infrastructure investment continuation into 2026. Founders should monitor these hyperscaler announcements for industry buying patterns and compute capacity constraints that directly impact AI startup opportunities and funding availability.
•Self-Driving Data Strategy: Tesla uses fully synthetic video generation to train FSD systems, creating unlimited edge case scenarios like deer crossings without real-world capture. Wave and Waymo employ similar approaches, suggesting synthetic data generation becomes standard methodology for autonomous vehicle development and safety validation.
•Expense Fraud Prevention: AI-generated fake receipts create new fraud vectors, but companies like Ramp deploy AI detection systems in response. Founders should implement stipend-based expense systems ($50-75 daily) rather than receipt tracking to eliminate fraud incentives and reduce administrative overhead while maintaining employee trust.
•Marketplace Revenue Reporting: Companies like Mercor paying 60-70% of gross revenue to service providers face scrutiny over ARR calculations. Founders must distinguish between gross marketplace volume and net retained revenue when reporting metrics to investors, as conflating these figures creates misleading valuation multiples and funding expectations.

Notable Moment

Calacanis reveals his early laser printer repair experience where coworkers deliberately slowed work to maximize billable hours, teaching him that small-scale dishonesty like fake receipts damages character more than large-scale theft because it reflects poorly on personal integrity.

Know someone who'd find this useful?