Meet the Authors

Key Takeaways What you need to know
  1. Perplexity’s finance AI agent connects licensed Morningstar, PitchBook, Daloopa, and Carbon Arc data while offering 35 prebuilt financial-analysis workflows.

  2. Source traceability links each reported figure to its filing, transcript, market-data source, or underlying calculation, giving SAP Finance teams a clearer audit trail.

  3. SAP S/4HANA Finance and Group Reporting teams must assess data access, governance, and audit logging before agent-generated analysis enters close and reporting workflows.

Perplexity launched Computer for Professional Finance, a finance-specific version of its agent aimed at corporate finance teams, analysts, and researchers. Built on four stated principles — easy access to trusted data, prebuilt workflows, full traceability, and portability across existing tools — it reframes Computer from a general assistant into something closer to a finance analyst that drafts, sources, and shows its work.

For SAP S/4HANA Finance and Group Reporting teams, the relevant question is narrower: Can an agent turn ledger and market data into a board-ready variance analysis that auditors can trace? Perplexity says it can. The audited close will test that.

Licensed Data Sits Beside Off-the-Shelf Tools

Perplexity describes two ways to get trusted data into the agent. Teams that already pay for Morningstar, PitchBook, Daloopa, or Carbon Arc can connect those subscriptions through MCP connectors, a standard interface that lets the agent authenticate into a provider a firm already licenses, so external benchmarks sit beside its own numbers without re-licensing.

Explore related questions

Alongside that, Perplexity says Computer ships with built-in tools drawing on data from 14 providers. For SAP Finance, the practical read is that market and reference data can sit next to internal actuals in one workflow.

Where SAP S/4HANA data is replicated into Snowflake for analytics, Perplexity’s separate Snowflake connector offers one path to reach that internal data, though that is a distinct enterprise release, not part of the finance product itself.

Prebuilt Workflows Target Repetitive Finance Work

The finance product ships with 35 dedicated workflows across segments including private equity, wealth management, and investment banking, covering recurring tasks. The pitch is that analysts start from a prebuilt workflow.

Those tasks rhyme with the work SAP Group Reporting teams repeat each cycle when they reconcile actuals against plan. The closer an agent’s prebuilt routines map to that recurring work, the more plausible it becomes inside an existing reporting process.

Source Traceability Is the Governance Story

Traceability is the principle SAP Finance teams should weigh most closely.

Perplexity says every numeric value links back to its source: hover any figure and the agent shows the SEC filing, earnings transcript, market-data page, or licensed source it came from, along with any calculations layered on top.

That maps directly onto the demands of a regulated close, where figures must trace to its origin and auditors expect to follow the math without reconstructing it. Exposing sources by default addresses a recurring concern about using generative tools in audited finance.

What This Means for SAPinsiders

  • Agentic AI moves into the finance close. SAP Finance teams are starting to test finance-specific agents, rather than general-purpose assistants, against existing reporting workflows.
  • Source traceability becomes table stakes. Regulated close and Group Reporting cycles require every figure to link back to its filing or warehouse origin — and teams are weighing tools on whether outputs hold up to that standard.
  • AI agents embed inside existing tools. With an Excel add-in and Microsoft Teams access, Perplexity is placing the agent inside the applications finance teams already use, putting data governance, access controls, and audit logging in scope.

Events

29Oct
SAPinsider Summit New Orleans 2026New Orleans, Louisiana, United States
View All