Skip to header Skip to main navigation Skip to main content Skip to footer

Dropbox Bets 350 Million Dollars on Agent Storage With Its Biggest Deal Ever

Breadcrumb

  • Home
  • Dropbox Bets 350 Million Dollars on Agent Storage With Its Biggest Deal Ever
A Dropbox logo installation on a sidewalk in San Francisco, with City Hall in the background.

Dropbox announced on Thursday, November 5, that it has agreed to acquire Tigris Data for about 350 million dollars in cash, in what the company says is its largest acquisition since it was founded in 2007. The definitive agreement was signed on November 3, the transaction is expected to close in the fourth quarter of 2026 subject to customary conditions, and Dropbox told TechCrunch it has no plans to reduce headcount as part of the deal.

The purchase is the engine behind DBX Agent Cloud, a new storage platform for businesses and developers building AI models and agents. It will launch with object storage sold as a standalone product, compatible with the tools and applications built for Amazon S3, so teams can buy capacity without moving into the rest of Dropbox. Planned additions go further: content processing that turns documents, images, audio, and video into material AI systems can use, and filesystem features that let agents manage files and keep continuity across tasks and sessions.

Tigris is a small company with an outsized fit for that job. Founded in 2021 in Sunnyvale, California, by former Uber engineers who helped build that company's storage infrastructure, it has raised about 31.9 million dollars in total, including a 25 million dollar Series A in October 2025. Its platform stores data across regions and positions it close to the users and compute that need it, which is exactly the problem agent workloads create: many small reads and writes, spread across clouds, with latency that compounds at every step of a long task. Dropbox chief business officer Eric Webster told Crunchbase News that Tigris brings a production product, an experienced team, and established relationships with AI developers.

The deal is one half of a two-part strategy. For existing customers, Dropbox is adding AI actions to its core product: users can ask AI to perform supported tasks on their files, choose between models, and work with Dropbox content inside tools such as ChatGPT, Claude, and Muse. The company says the updated experience has already reached more than 70,000 teams, starting with eligible Standard and Advanced customers in the United States, the United Kingdom, and Australia, with broader availability planned for 2027. DBX Agent Cloud, by contrast, is aimed at people who may never open the Dropbox app at all.

The timing deserves notice, because it runs against the week's other enterprise signal. Bloomberg reported on November 4, via Reuters, that Apple cut its engineers' daily AI spending allowance from 300 dollars to 200 dollars in October, is considering 100 dollars for some teams, and may take some non-engineering limits as low as 25 dollars. One company is rationing tokens while another spends 350 million dollars on the storage layer underneath them. Both are really making the same point: the cost of agents is no longer the model bill alone. Builders who treat storage, retrieval, and file continuity as afterthoughts will pay for it in latency and rework, and a portable, S3-compatible layer that is not tied to one cloud is a credible place to start fixing that.

Rocky Mountain Madman LLC
Enterprise AI
AI Product News
AI Implementation
Web Services
AI Funding

Ai Industry News

Nvidia's Nemotron 3.5 Transcribe Puts 40 Languages in One Open-Weight Model

Nvidia released Nemotron 3.5 Transcribe on Wednesday, a free, open-weight speech model that folds transcription, speaker labels, translation, and spoken question answering into a single promptable checkpoint. It transcribes audio in up to 40 languages, translates 31 of them to and from English, and can tell up to eight speakers apart in recordings up to an hour long.

  • Read more about Nvidia's Nemotron 3.5 Transcribe Puts 40 Languages in One Open-Weight Model

Google's Genie 4 Gives Its World Model a Voice, and Lets You Rewrite the Scene Live

Google DeepMind announced Genie 4 on October 27 as a research preview for trusted testers, and the headline change is sound. The new world model generates explorable, interactive scenes in real time from a text prompt or an image, and for the first time it produces audio natively alongside the picture: ambient noise, music, and characters who can converse and react as events unfold. DeepMind says scenes now hold together for up to about two minutes, roughly double the limit reported for Genie 3, and every video and audio clip it emits carries a SynthID watermark.

  • Read more about Google's Genie 4 Gives Its World Model a Voice, and Lets You Rewrite the Scene Live

OpenAI Says Its Models Disproved a Core Complexity Conjecture. Now Experts Have to Check the Proof

A mathematician writing a proof in chalk on a blackboard covered with equations.

OpenAI announced on October 26 that its internal models have produced three new scientific results: a proof of Sogge's local smoothing conjecture in harmonic analysis, a disproof of the Exponential Time Hypothesis in computer science, and a rigorous existence proof for quantum chromodynamics and related quantum field theories in physics. None of the three has been peer reviewed. OpenAI is unusually candid about that, calling the manuscripts raw research data and handing them to outside experts to examine, explain, and, if they hold up, turn into actual scientific understanding.

  • Read more about OpenAI Says Its Models Disproved a Core Complexity Conjecture. Now Experts Have to Check the Proof

Pagination

  • First page « First
  • Previous page ‹ Previous
  • …
  • Page 5
  • Page 6
  • Page 7
  • Page 8
  • Current page 9
  • Page 10
  • Page 11
  • Page 12
  • Page 13
  • …
  • Next page Next ›
  • Last page Last »

Copyright © 2026 Rocky Mountain Madman LLC - All rights reserved

Developed and Designed by Rocky Mountain Madman LLC