AfterQuery's $3.2B, Felix's $200M WhatsApp Play & Sequoia's Empirik: AI Startups to Watch for September 2, 2026

The AI startup funding machine keeps humming. Here are the AI startups pulling in the biggest rounds today. AfterQuery became Y Combinator's fastest-ever unicorn at $3.2 billion, just five months after its $300 million Series A. Felix raised $200 million to turn WhatsApp into a financial platform for Latino immigrants. Sequoia spun out an AI infrastructure startup focused on preventing outages before they happen. And Adobe quietly picked up an Indian marketing intelligence team.

Today's Top AI Startups: AfterQuery, Felix, and Empirik

See yesterday's AI Startups to Watch for more moves from the last week.

AfterQuery: YC's Fastest Unicorn at $3.2 Billion

AfterQuery, an AI training-data startup founded by two 22 and 23-year-old entrepreneurs, has reportedly raised a round that valued it at $3.2 billion. That is more than a 10x increase from its $300 million valuation in April, making it the fastest startup in Y Combinator history to go from launch to unicorn status.

The San Francisco-based company reached a $100 million annualized revenue run rate by April and works with Nvidia, Legora, and Korean AI lab Motif Technologies. AfterQuery employs knowledge professionals like doctors, lawyers, and specialists to train AI models not just on correct answers, but on how experts actually work: "encoding the patterns, decisions, and reasoning of the world's best practitioners."

The founders attended Y Combinator's Winter 2025 cohort, just 18 months ago. Forbes first reported on the round.

Sources: TechCrunch, Forbes

Felix Raises $200M to Build a Financial Platform on WhatsApp

Felix, a Miami-based AI-powered remittance platform for Latino immigrants, closed a $200 million Series C co-led by Andreessen Horowitz and General Catalyst. A16z led $87 million in equity, with QED Investors, Castle Island Ventures, and others participating.

Felix started with remittances, identifying sending money home as one of the most common financial needs among its users. The new funding aims to expand beyond remittances into a broader AI-powered financial platform. The company uses WhatsApp as its primary interface, reaching users where they already communicate rather than asking them to download a separate app.

The round signals continued investor confidence in vertical fintech platforms that serve specific immigrant communities with culturally tailored products.

Sources: Crunchbase News, Techmeme

Sequoia Spins Out Empirik: AI That Prevents Outages Before They Happen

Empirik officially launched as an independent company on September 1, raising $21 million in seed funding from Sequoia Capital, Canapi Ventures, and Alumni Ventures. The startup spun out of a two-year incubation at Sequoia to build what it calls an autonomous infrastructure engineer.

The problem Empirik targets is real and growing. Gartner predicts 90% of organizations will experience an AI-caused outage by 2029, even as they use AI tools to speed up development. Empirik tracks every system change across large infrastructure environments and models how changes ripple through dependencies.

CEO Kartik Chandrayana frames it simply: "What agentic AI did for software, Empirik wants to do for infrastructure engineering." Early customers include S&P Global and Guardant Health.

Sources: TechCrunch, TechTimes

Adobe Acquires Indian Startup Rilo

Adobe has acquired India-based marketing intelligence startup Rilo in a deal involving licensing and team acquisition. This is Adobe's second acquisition from India after buying video platform Rephrase.ai in 2023.

Founded in 2025 by IIT batchmates Georgi Boby and Dhruv Jaglan, Rilo raised $1 million at a $10 million valuation from Peak XV, DeVC, and Day Zero Ventures. The six-person team built workflow automation for go-to-market teams, including competitor intelligence, content repurposing, and sales call analysis. Rilo will shut down as a standalone product.

The acquisition fits Adobe's broader strategy of bolting AI-powered marketing tools onto its creative suite, following the $1.9 billion Semrush acquisition last year.

Sources: TechCrunch

Wafer Raises $40M, Turns Down Acquisition Offers

Wafer, an AI inference startup that uses autonomous agents to optimize GPU performance, raised $40 million in Series A funding at a valuation above $200 million. The company turned down multiple acquisition offers from cloud and inference providers to stay independent.

Based in San Francisco and backed by Y Combinator, Wafer went from zero to $8 million in ARR in four months after launching its cloud. The startup uses non-Nvidia chips and agent-based optimization to squeeze more performance out of existing GPU infrastructure, a value proposition that clearly resonated with acquirers.

Sources: The Information, Crypto Briefing

The thread connecting today's startup stories is clear: the money is flowing toward the infrastructure layer, not just the model layer. Companies building the tools to make AI work in production are pulling in the biggest rounds and the fastest valuations.

Frequently Asked Questions

How does AfterQuery compare to Scale AI and Mercor in the AI data market?

AfterQuery, Scale AI, and Mercor all provide human-vetted training data to frontier AI labs, but with different approaches. Scale AI pioneered the category and is now in talks with Nvidia at a $20 billion valuation. Mercor raised at $10 billion in 2025 and recruits skilled workers via AI interviews. AfterQuery's edge is custom validation software that screens expert-generated data within a 'Goldilocks' difficulty range, plus it publishes research proving its data improves benchmark scores.

What is AfterQuery's actual revenue and how does it scale?

AfterQuery surpassed $100 million in annualized revenue by April 2026 when it raised its Series A. Founder Spencer Mateega later posted on X that recurring revenue had reached the 'hundreds of millions' — a significant jump in just months. The company runs its own internal post-training pipeline, training models on its data first so labs can verify quality before purchasing. This proof-of-value approach has attracted customers including Nvidia, Legora, and Korean AI lab Motif Technologies.

What role did AfterQuery play in Nvidia's Nemotron models?

AfterQuery is the only data partner named in Nvidia's Nemotron 3 Ultra technical report. Nvidia used AfterQuery's Off-The-Shelf Office Agent Training Dataset to improve Nemotron's GDPval performance, applying the data in two stages: light SFT on the student model and pivot RL. Nemotron 3 Ultra is a fully open 550B-A55B LatentMoE model with open weights and training data. AfterQuery's work contributed to Nemotron achieving competitive scores on professional workflow benchmarks.

Related reading

Related reading from earlier coverage, with the numbers behind each one:

• orbital data centers, a chipmaker taking on Nvidia, and a Sequoia-backed security model: AI Startups to Watch for August 22, 2026
• Starcloud's .3B orbital data centers, a 1B inference chip bet, and robot models that learn from a single demo: AI Startups to Watch for August 23, 2026
• HUMAIN's Saudi Robot Trucks, Google's $40M Hollywood IP Bids & UK's £100M AI Fund: AI Startups to Watch for August 31, 2026
• Perplexity, Tencent, Slack & fal Ship Agent Tools: AI Startups to Watch for August 30, 2026