Augment Code Is Selling Pure Storage's 2,000 Engineers an Agent That Reads Their Whole Repo

The Palo Alto company has raised $252M from Sutter Hill and Meritech to chase enterprise codebases that Cursor and Copilot still struggle to load.

About Augment

Published

When Pure Storage put roughly 2,000 engineers on Augment Code, the pitch was not faster autocomplete. It was an AI agent that could hold the entire codebase in context, follow dependencies across services, and answer questions a new hire would take a quarter to figure out [Cerebral Valley, 2026]. That deployment, alongside MoneyGram, AMP Robotics, Newfront Insurance, and Peek Travel [Augment Code; Cerebral Valley, 2026], is the story Augment is telling buyers right now: not a smarter editor, but a coworker that has actually read the repo.

Founded in Palo Alto in 2022 by Guy Gur-Ari, Igor Ostrovsky, and Scott Dietzen [Crunchbase; Startup Intros], Augment Code calls itself "The Software Agent Company" [Augment Code]. The product surface now spans an IDE assistant, a CLI, and an AI code reviewer that plugs into GitHub pull requests with hooks into Jira and Notion [SiliconANGLE, 2025; Cerebral Valley, 2026]. Enterprise plans remove seat limits on Code Review across an organization [Augment Code], which shifts the negotiation from per-developer math to a platform line item.

The bet

The wedge is context. Cursor, GitHub Copilot, and Windsurf [SiliconANGLE, 2025] all compete for the developer's editor. Augment's argument is that real enterprise codebases break the assumptions that consumer-grade coding tools were built on. The company's marketing leans on a "context engine" that indexes the whole repo and grounds agent suggestions in it [Augment Code]. If that claim holds up under load at customers like Pure Storage, the ICP becomes clear: engineering organizations of 500 or more developers working on long-lived, proprietary codebases.

Why it could be big

The market tailwind is real. Every Fortune 500 CIO is being asked by the board what the AI coding strategy is. Sutter Hill Ventures, which led an early round, and Meritech Capital Partners [Startup Intros; Crunchbase] are investors who tend to back enterprise-software companies with long sales cycles and durable contract value. The $252M raised across rounds [Startup Intros] gives Augment the runway to do the unglamorous work that enterprise software requires: SOC 2, SSO integrations, and a field engineering team that can sit with a customer during rollout.

Metric Value
Total disclosed funding $252M
Pure Storage engineers on platform 2,000 engineers
Years since founding 3 years

The team and traction

Scott Dietzen brings the most legible enterprise pedigree of the founding group: he was previously CEO of Pure Storage, the same company now cited as a flagship Augment customer [Cerebral Valley, 2026]. Igor Ostrovsky's background in distributed systems and Guy Gur-Ari's research profile round out a team that is being asked to sell into the most technical buyer in the building. Customer logos disclosed so far (Pure Storage, MoneyGram, AMP Robotics, Newfront Insurance, Peek Travel) [Cerebral Valley, 2026; Augment Code] span storage infrastructure, payments, robotics, insurance, and travel.

The honest counterfactual

The competitive set is the part a buyer should weigh hardest. Cursor has captured enormous developer mindshare and is reportedly raising at valuations that dwarf Augment's disclosed total funding. GitHub Copilot ships inside an enterprise agreement most large companies already have. The bear case is that coding assistants commoditize toward the model layer, and whichever vendor has the cheapest distribution wins. Augment's answer is that the differentiation is not the model but the retrieval and context layer wrapped around the customer's own code.

What to watch

Three things to track over the next year. First, whether Augment discloses a second flagship customer at Pure Storage's scale. Second, whether the company ships a deployment option (VPC, on-prem, or bring-your-own-model) that unblocks regulated buyers. Third, the next round: at $252M raised and a Series B label [Startup Intros], a Series C in 2025 or 2026 would put a hard market price on the context-engine thesis.

Read on Startuply.vc