Recommended Articles
Tech & AI

Meta Strategy Chief Dave Wehner Leaves to Become CFO of AI Startup Fireworks

By Angeled · November 7, 2026 · Tech & AI
Meta Strategy Chief Dave Wehner Leaves to Become CFO of AI Startup Fireworks
Meta Strategy Chief Dave Wehner Leaves to Become CFO of AI Startup Fireworks

Meta chief strategy officer Dave Wehner is leaving the company to become chief financial officer of Fireworks, an AI inference startup. The move was reported by The Information on 6–7 November 2026, and Wehner confirmed on Friday that he will make the switch.

Wehner joined Facebook in 2012 as vice president of corporate finance and business planning, was promoted to chief financial officer in 2014, held that role for more than eight years, and moved to chief strategy officer in 2022. His departure ends a 14-year tenure at the company, spanning its transformation from a social network into one of the world's biggest AI spenders.

Who is Dave Wehner?

Wehner is one of the most experienced finance executives in big tech. As Meta's CFO from 2014 to 2022, he oversaw the company's finances through the Instagram and WhatsApp integrations, the shift to mobile advertising, the multibillion-dollar Reality Labs investment, and the 2021 rebrand from Facebook to Meta.

As chief strategy officer since 2022, he has worked on Meta's longer-term strategic questions, including its AI strategy, at a time when the company has committed tens of billions of dollars a year to AI infrastructure. Losing a strategist with that depth of institutional knowledge is a significant departure for Meta, even for a company accustomed to executive turnover.

What is Fireworks?

Fireworks is an AI inference company founded in 2022 by Lin Qiao and six co-founders, most of whom had worked on PyTorch — the open-source AI framework — at Meta and Google. Qiao herself previously led PyTorch at Meta, where she managed more than 300 engineers. Fireworks is based in Redwood City, California.

To understand Fireworks, it helps to understand the difference between training and inference. Training is the expensive, one-off process of building an AI model; inference is running that model to answer real user requests, millions or billions of times a day. As AI moves from research labs into products used by hundreds of millions of people, inference becomes the larger and more commercially important market — and the one where speed and cost per request decide who wins.

Fireworks runs open-weight and custom large language models on its own cloud infrastructure and charges enterprises per token, per GPU hour or under reserved-capacity contracts. Its pitch is that it can serve models faster and cheaper than the big cloud providers, using proprietary optimisation technology, and that customers can fine-tune models on their own data. Its customers include companies such as Uber, Shopify, DoorDash and Cursor, and it says it processes more than 40 trillion tokens a day.

The company's rise has been extraordinarily fast. It raised a $52 million Series B in July 2024 at a $552 million valuation, a $250 million Series C in October 2025 at roughly a $4 billion valuation, and a $1.505 billion Series D in July 2026 at a $17.5 billion valuation, led by Atreides Management, Index Ventures and TCV, with Nvidia participating. It reported crossing $1 billion in annualised revenue by July 2026 — about five times its level a year earlier.

Why a CFO hire like this matters

Hiring a CFO of Wehner's calibre is a signal in itself. Startups typically hire a big-company CFO when they are preparing for the disciplines of public-company life: audited financials, investor relations, a possible IPO, or simply managing capital at a scale where mistakes are expensive. Fireworks, at a $17.5 billion valuation with over $1 billion in annualised revenue and a possible new round at $30–40 billion, is operating at a scale where world-class financial leadership is no longer optional.

For Wehner, the move is also a bet on where value will be created in AI. He could have stayed at Meta, one of the richest companies in the world. Instead he is joining a 2022-founded startup — a decision that suggests he believes inference infrastructure, not just model-building, is where the next great AI company will be built. It is also a homecoming of sorts: Fireworks was founded by his former Meta colleagues from the PyTorch team, so he is joining people whose work he knows well.

The wider talent war

Wehner's departure fits a broader pattern: senior executives and researchers are leaving the largest tech companies to join — or found — AI infrastructure startups, where the upside is larger and the work is more focused. The flow used to run mostly in one direction, from startups into big tech; in AI infrastructure, it now runs both ways.

For Meta, the departure raises a practical question about who owns strategy at a moment when the company is making some of the largest capital commitments in its history. For founders, the lesson is about hiring: the best CFO for a company approaching public scale is often someone who has already done the job at that scale, and those people are increasingly willing to join startups if the mission and the equity are compelling.

For investors, Wehner's move is a data point worth weighing alongside the financials. When a veteran who has seen Meta's books from the inside chooses an inference startup over staying, it suggests the inference market is bigger and more durable than the sceptics argue — and that the competition among inference providers is about to get much more serious.

Angeled.com connects founders with investors.

Read Next