Perplexity, the San Francisco-based AI company best known for taking on Google in the search space, has quietly been working on something far more ambitious — and on March 12, 2026, it made that ambition official. The company launched Computer for Enterprise, bringing its powerful multi-step agentic AI system into the hands of business teams across the globe.

If you’ve been following Perplexity’s trajectory over the past couple of years, none of this should come as a complete surprise. The startup, founded by Chennai-born CEO Aravind Srinivas, has grown from a scrappy search challenger into a $20 billion company, with backers including NVIDIA, Jeff Bezos, and SoftBank Vision Fund 2. That kind of valuation doesn’t come easy — and it certainly doesn’t hold up if all you’re doing is returning better search results. Perplexity clearly knows this, and Computer for Enterprise feels like its most direct answer yet to the question every investor has been asking: now what?​

The roots of the product go back to late February 2026, when Perplexity first introduced Perplexity Computer to its Max subscribers — a $200-per-month tier for power users. That version was already strikingly capable: a cloud-based agentic system capable of orchestrating 20 AI models, including Claude, Gemini, and others, to handle complex, multi-stage workflows without the user having to lift a finger. Think of it less like a chatbot and more like a digital coworker — one that doesn’t sleep, doesn’t miss deadlines, and doesn’t ask for vacation days. Now, that same technology has been tuned and packaged specifically for enterprise environments, and it’s available to Enterprise customers today.

What makes Computer for Enterprise genuinely interesting isn’t just the breadth of what it can do — it’s how seamlessly it’s designed to fit into the tools companies are already using. Perplexity hasn’t asked teams to overhaul their software stack. Instead, the product connects to Salesforce, Microsoft Teams, HubSpot, MySQL, GitHub, and more than 400 other tools through app connectors, pulling in live company context as teams work. This is a smart play. Enterprise software adoption often fails not because the technology is bad, but because the switching costs are too high. Perplexity is essentially saying: keep using what you have — Computer will come to you.

The use cases span industries in ways that make it easy to see why enterprise teams would be excited. Finance teams, for instance, can build an interactive due diligence tracker that analyzes M&A documents and flags risks from a data room automatically. Legal teams can review a vendor agreement, compare versions, and redline a contract — all within a single workflow, without bouncing between tools. Marketers can create entire campaigns, from ad creative to landing pages, and then ask Computer to spin up a performance dashboard to track results in real time. These aren’t hypothetical demos dreamed up for a press release — they’re workflows that Perplexity’s own internal teams have already put to use.​

And the internal results have been nothing short of remarkable. In a study of over 16,000 queries, Perplexity found that Computer saved its own teams $1.6 million in labor costs, and completed the equivalent of 3.25 years of work in just four weeks. Those numbers are striking, and will inevitably raise eyebrows among skeptics — but the company says the data comes from real internal usage, not a controlled pilot. For enterprise buyers who need to justify AI spend to CFOs and boards, that kind of ROI framing is going to be very hard to ignore.​

It’s worth zooming out a little to appreciate just how significant a shift this represents for Perplexity as a company. When Srinivas launched the product back in 2022, the pitch was simple: a better way to search the internet, with citations. That was a valuable product, and it found a real audience. But the AI landscape has moved at extraordinary speed. Rivals like OpenAI, Anthropic, and Google have all pushed aggressively into agentic territory, and the race now isn’t just about who can answer a question fastest — it’s about who can do the work. Perplexity’s pivot from “knowledge engine” to “action engine,” as observers have described it, is a direct response to that shift.

The architecture behind Computer for Enterprise reflects this thinking in detail. Rather than betting on a single frontier model, Perplexity has built the system around orchestration — intelligently routing tasks across 20 different models, choosing the right one for each job. This model-agnostic approach is an acknowledgment that no single AI system is best at everything, and that the real competitive advantage lies in knowing how to stitch them together. Forbes noted that this kind of multi-model pipeline strategy is actually mirroring what many sophisticated enterprises are building internally — Perplexity is just packaging it up and making it accessible.

Beyond broad task handling, the enterprise edition also introduces a feature called Skills, which lets teams give Computer preset instructions tailored to their specific business processes. A team could, for example, create a Skill that tells Computer exactly how their organization structures a product requirements document, or how to research and personalize outreach to a new sales prospect. It’s a layer of customization that transforms Computer from a general-purpose tool into something that actually understands how a particular company operates.​

One area where Perplexity has been particularly careful is security — a non-negotiable for enterprise buyers. Computer for Enterprise is built on the company’s existing enterprise foundation, which is already SOC 2 Type II certified. Every task runs in an isolated cloud environment, complete with authenticated integrations and secure browser and code execution inside a sandbox. SAML SSO, audit logs, and administrative controls round out the package. In a world where enterprises are increasingly worried about data leakage and AI governance, that level of built-in compliance infrastructure is the kind of detail that gets deals signed.

The product also has a notably natural home in Slack, which is where it was originally developed as an internal tool. Teams can DM Computer directly, or bring it into a shared channel and work alongside it in the flow of conversation — no separate interface, no friction, just the assistant plugged into the workspace where work is already happening. For companies that have made Slack the connective tissue of their operations, this is a genuinely low-barrier entry point.​

Perplexity’s journey from a search startup to an enterprise AI platform has been fast, and there’s clearly a lot riding on Computer’s success. The company’s valuation climbed from $500 million to $20 billion in just 18 months, and investors are now watching closely to see whether that growth can be justified with real enterprise revenue. Computer for Enterprise — announced at the company’s Ask 2026 developer conference and now available to all Enterprise customers — looks like Perplexity’s most serious play yet to show that it isn’t just a smarter search engine, but a legitimate AI platform for how business actually gets done.