Perplexity Computer Expands to Microsoft Teams, Excel, and Professional Finance

Perplexity is moving fast on the enterprise front, and the latest round of updates to its Computer platform is the clearest signal yet that it's positioning for the center of professional workflows—not the edges.

In a single announcement: Computer is now available in Microsoft Teams, launching as a native side panel in Excel (beta), adding data connectors for Snowflake and Databricks, partnering with 1Password for credentialed agent security, launching a dedicated Professional Finance tier with licensed data from Morningstar, PitchBook, Daloopa, and Carbon Arc, and rolling out over 70 pre-built enterprise workflows. Oh, and there's a personal compute layer for Mac that runs 24/7.

That's a lot of surface area covered at once.

Microsoft Teams and Excel: Meeting Workers Where They Already Are

The Teams integration is straightforward in concept and significant in reach. Microsoft reports over 350 million monthly active Teams users worldwide. Computer can now be messaged directly or pulled into a channel—no context switching, no separate app. The same thread where a colleague prompted it last week becomes the starting point for the next person with a similar question.

The Excel integration is arguably more interesting. Analysts don't leave Excel—they live in it. A native side panel means Computer sits alongside the model already open, available for questions in place. That's a different posture than asking someone to open a separate tool. It meets a high-value user at the exact moment they need help, inside the artifact they're already building.

Workflows: Bundling Repetition Into Repeatable Starting Points

The 70+ workflow library addresses something real. In most organizations, a small number of task types account for a disproportionate share of AI requests—competitive analysis, call prep, deal sourcing, reporting. Perplexity's workflows bundle the prompt, context, and output format for a specific task into a single reusable starting point. Teams can share, customize, schedule, and run them asynchronously.

The scheduling piece is worth noting. Work that continues after the person who started it has moved on is a meaningful operational shift—closer to automated process than assisted task.

Data Connectors and Credentialed Security

The Snowflake and Databricks connectors address a friction point that has limited enterprise AI adoption more than most vendors acknowledge: the underlying data is almost always proprietary, credentialed, and siloed. Perplexity says teams can now query internal data without copying it between systems or routing every request through a data science team.

The 1Password partnership handles the credential problem on the agent side. Computer can act on behalf of a user without the underlying credentials touching the agent or the model—actions remain authorized, governed, and auditable. For enterprise security and compliance teams, that's a meaningful guardrail.

Professional Finance: Depth Over Generality

The finance tier is the most specific product bet in this announcement. General-purpose AI tools hit a ceiling in finance because deliverables require sourced, auditable figures tied to specific vendors. Perplexity is pulling from licensed data providers directly—tearsheets, annotated stock charts, and equity research comparisons where every figure links back to its source. It runs inside Excel alongside existing analyst workflows.

The implicit argument is that vertical depth beats horizontal generality for high-stakes professional work. That's a reasonable bet, and if it holds, other domains are likely next.

What Marketing and Growth Teams Should Take From This

The pattern Perplexity is executing on—meet users inside the tools they already use, bundle repeated tasks into reusable workflows, connect to proprietary data, make credentials auditable—is a template worth understanding regardless of whether you use Computer specifically.

The AI tools that will embed deepest into enterprise workflows won't be the most powerful in isolation. They'll be the ones that reduce the friction of getting from question to finished artifact inside the systems teams already depend on. For marketing operations specifically, that means the evaluation criteria for AI tooling should increasingly include integration depth and workflow repeatability, not just output quality.

The infrastructure for always-on, deeply integrated AI work is being built right now. The strategic question is whether your team's processes are designed to use it.

Building AI into your marketing workflows in a way that actually sticks requires more than picking the right tool. Winsome Marketing's growth team helps you design the strategy behind the stack. Let's talk.