AI Financial Platform Brex Inks 94K-SF Lease at GFP’s 200 Varick Street

Credited to Commercial Observer · commercialobserver.com · 1h ago

Brex, an AI-powered financial platform, is moving into a new office in Manhattan’s Hudson Square neighborhood.

The San Francisco-based artificial intelligence firm, which provides corporate cards, expense management software and business cash management accounts for new companies, has signed a lease for 93,779 square feet at GFP Real Estate ’s 200 Varick Street, according to a third-quarter Manhattan office report from CBRE.

The deal likely represents a relocation for Brex, which moved into its current New York City headquarters around January 2025. The exact address of Brex’s current office is unclear.

The length of Brex’s new lease at 200 Varick and the asking rent were also unclear, but CBRE’s report found office rents in Midtown South averaged $86.94 per square foot during the third quarter of 2026.

It’s also unclear who brokered the deal, but GFP has repped itself in-house in past deals at 200 Varick. GFP declined to comment, while a spokesperson for Brex did not immediately respond to a request for comment.

Brex was founded in 2017 by Henrique Dubugras and Pedro Franceschi as a way for early-stage startups without meager credit histories to secure corporate credit cards. Capital One acquired the firm in April for a final valuation of $5.15 billion.

Brex also notched one of the largest office leases signed in Midtown South during the third quarter of this year, following Anthropic ’s whopping 465,630-square-foot deal at 330 Hudson Street, the New York City Human Resources Administration ’s 174,186-square-foot renewal at 260 11th Avenue, and AI firm Legora ’s 98,420-square-foot lease at 11 Madison Avenue, according to CBRE.

Isabelle Durso can be reached at idurso@commercialobserver.com.

Models & LabsUnited States

Automation Anywhere acquires Boost.ai to expand customer-facing voice AI

Credited to SiliconANGLE · siliconangle.com2h ago

Useful

Automation Anywhere Inc., an enterprise agentic automation and orchestration firm, announced Wednesday an agreement to acquire Boost.ai Inc., a conversational voice artificial intelligence company, from Nordic Capital.

The firm seeks to build the “autonomous enterprise,” an operating model where businesses operate up to 80% autonomously or with AI assistance.

Boost.ai’s technology will extend that vision into customer experience, allowing AI to take requests from customers or employees and guide them through the middle and back office.

“Every enterprise runs on three kinds of work: serving customers, supporting employees, and running the business. Most run in isolation,” said Automation Anywhere Chief Executive Mihir Shukla. “With Automation Anywhere and Boost.ai, we intend to be the leading offering to give every enterprise an autonomous operation that listens, decides and acts across all three.”

Currently, Boost.ai serves hundreds of customer organizations, with a particularly strong base in regulated industries and Europe. Many of its customers focus on financial services, telecommunications, and insurance. This will expand Automation Anywhere’s thousands of global customers who operate in similarly complex industries.

Boost.ai’s platform supports more than 36 languages and ensures message security and privacy, with compliance with Europe’s General Data Protection Regulation and the Health Insurance Portability and Accountability Act.

“Boost.ai’s deep expertise in regulated industries has proven that conversational and voice AI can operate at enterprise scale with the reliability, compliance, and governance that regulated industries require,” said Boost.ai CEO Jerry Haywood.

This acquisition follows Automation Anywhere’s acquisition of Aisera Inc. in late 2025, integrating its agentic AI platform into its agentic process automation ecosystem. The company said combining the two companies’ platforms will give Automation Anywhere the most complete offering on the market for autonomous work across office work.

Original · SiliconANGLE

Useful

View the articleSiliconANGLE

Front Method