Dropbox buys Tigris Data for $350 million to enter AI infrastructure

The deal: $350 million for Dropbox's biggest acquisition ever

Dropbox announced on Thursday (5th) plans to acquire Tigris Data for about $350 million, according to Crunchbase News. The company confirmed to the outlet that this will be its biggest acquisition since its founding, in 2007. The deal is still in the planning stage, not finalized.

Tigris builds a network of localized storage data centers, designed to serve the distributed workloads of modern AI. According to Dropbox, the purchase delivers all at once a product already in production, an experienced team, and relationships already established with AI developers, a faster shortcut than building all of that in-house.

To put the value in perspective, here's how Tigris compares to Dropbox's other major purchases over the years:

Who is Tigris Data

Founded in November 2021 by former Uber engineers, Tigris is based in Sunnyvale, California. The startup has raised about $31.9 million since its founding, according to Crunchbase data. The Series A, closed in October 2025, brought in $25 million led by Spark Capital, with participation from Andreessen Horowitz and Basis Set Ventures, which had also led the seed round. General Catalyst is among the earlier investors.

At the time of the Series A, CEO and co-founder Ovais Tariq told TechCrunch that the company had been growing 8x year over year since its founding. What Dropbox is effectively buying is a storage platform built for developers whose applications run across multiple cloud providers at once, not locked into a single vendor.

DBX Agent Cloud: the bet on object storage

Alongside the acquisition announcement, Dropbox unveiled DBX Agent Cloud, which will initially offer standalone object storage: a way to store and retrieve large collections of data, including images, video, and the files used to train and run AI models. It's a different audience from the end user who today syncs folders and shares files through Dropbox.

Eric Webster, chief business officer at Dropbox, explained to Crunchbase News why the acquisition speeds up this entry:

Tigris brings together a product already running in production, an experienced team, and established relationships with AI developers. That combination would accelerate our entry into infrastructure for AI applications through DBX Agent Cloud.

Eric Webster, chief business officer at Dropbox

Two AI paths within the same company

Dropbox's AI strategy runs on two separate fronts. For existing product users, the company is bringing AI directly to existing files: searching, understanding, creating, editing, and organizing content, with the ability to switch between supported models and work with Dropbox files in tools like ChatGPT and Claude, with results flowing back into the platform.

For businesses and developers, DBX Agent Cloud extends Dropbox's storage capacity to the infrastructure behind AI applications and agents, a new market for the company. Webster sums up the division of roles: Agent Cloud opens up a developer and infrastructure market, while AI inside Dropbox exists to deepen engagement, retention, and expansion among customers already there.

A slower pace of acquisitions, but bigger bets

Dropbox has made 35 known acquisitions since its founding, according to Crunchbase, but the pace has changed considerably. Between 2012 and 2015, the company made 23 purchases, including nine in 2014 alone. By contrast, 2024 and 2025 added up to just four acquisitions, an average of two per year, about 65% below the pace of its early phase.

The most recent purchases show a deliberate focus: Reclaim.ai (scheduling) and Nira (document security) in 2024, followed by Promoted.ai and Mobius Labs in 2025. AI productivity, search, and security had been topping the list so far, and Tigris marks a shift toward pure infrastructure.

What changes for those building on top of this

The takeaway for developers is more about platform choice than about a new feature in the Dropbox you already use. Tigris was built to serve teams that distribute AI workloads across multiple cloud providers; the acquisition places that proposition inside a publicly traded company, with scale and cash reserves far larger than a five-year-old startup's.

This raises some concrete questions for anyone already running data on top of Tigris or evaluating where to store training datasets, embeddings, and model artifacts:

• Product continuity: Dropbox describes Tigris as something already in production, which suggests intent to keep the service running, but post-acquisition pricing and contract migrations are the rule, not the exception.
• Multicloud flexibility: part of Tigris's value was not locking developers into a single provider; it's worth watching whether that promise survives inside a Dropbox that now wants to sell its own AI stack.
• New competitor in object storage: DBX Agent Cloud enters a market currently dominated by S3, Google Cloud Storage, and Azure Blob, specifically competing for the niche of AI training and inference data.

The broader signal is one of consolidation: cash-rich incumbents prefer to buy AI infrastructure startups already validated in production rather than build from scratch, which accelerates the acquiring company's roadmap but also concentrates architecture decisions that used to be spread across smaller, more specialized vendors. For anyone choosing where to host AI data today, Agent Cloud still has no pricing, SLA, or public documentation: the acquisition has been announced, not finalized, and the offering's technical details should only appear after the deal closes.