Perplexity Computer adds finance workflows

// 45d agoPRODUCT UPDATE

Perplexity Computer adds finance workflows

Perplexity is expanding Computer into a finance-focused workflow layer for professional research teams. The update adds licensed data connectors for sources like Morningstar, PitchBook, Daloopa, and Carbon Arc, plus integrated market-data tools, so analysts can move from research to finished outputs such as tear sheets, memos, dashboards, and annotated charts with source traceability preserved end to end.

// ANALYSIS

Hot take: this is Perplexity trying to compress the Bloomberg Terminal workflow into an AI-native research and production system, not just an answer engine.

–The key differentiator is not raw model quality; it is the combination of licensed data, workflow templates, and output formatting.
–Finance users care about provenance, and Perplexity is leaning hard into traceability at the point of every number and chart.
–The product’s value compounds if teams actually adopt the saved workflows instead of treating it like another chat surface.
–The risk is execution trust: finance teams will test whether the connectors, calculations, and citations hold up under real diligence.

// TAGS

aifinanceresearchworkflowsdata-connectorsagententerprise

DISCOVERED

45d ago

2026-05-06

PUBLISHED

45d ago

2026-05-06

RELEVANCE

8/ 10

AUTHOR

WorldofAI

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

RESEARCH29m ago

LLM Brewing maps code reasoning layers

LLM Brewing is the code artifact for a June 16 arXiv paper tracing when code-reasoning answers become readable in LLM hidden states, when models can actually decode them, and why some answers degrade later. The project ships benchmark, probing, CSD, diagnostic, and causal-validation pipelines on a separate experiment branch.

RESEARCH29m ago

LoopWM loops transformers for world models

LoopWM is a new arXiv research paper proposing looped transformer depth for world models, using a shared recurrent block to iteratively refine latent environment states. The authors claim up to 100x parameter efficiency, adaptive compute per transition, and stronger long-horizon prediction on ScienceWorld-style simulation tasks.

NEWS30m ago

Mercury Agent eyes GLM-5.2 default

Mercury Agent says GLM-5.2 is performing well enough to consider making Z.ai's model its default provider. The move would matter because Mercury is an open-source, always-on coding agent built around permissioned tools, memory, skills, and multi-channel operation.