How Perplexity built enterprise-grade financial research capabilities with Quartr API
When AI search enters financial research, the margin for error disappears. A misattributed quote, a delayed transcript, or incomplete coverage doesn't just create a bad user experience. It immediately breaks trust with professionals who constantly need to make high-stakes decisions.
For Perplexity, building credible earnings research capabilities meant solving a problem that has challenged financial platforms for years: how do you deliver institutional-quality data with consumer-grade speed and simplicity?
The answer required finding a data partner that could deliver first-party earnings content – live transcripts, audio, and company disclosures – without coverage gaps or reliability issues.
Breaking the institutional research divide
In a conversation with Jeff Grimes, Head of Live Events at Perplexity Finance, he outlined the market gap the team set out to fill and how they did it.
"Financial research has fallen into two buckets to date: beneficial to the retail investor, or at a higher price point, beneficial to the institutional investor."
Retail investors use the Perplexity Finance dashboard to quickly understand what's happening in markets and find information about assets: historical price data, fundamentals, analyst estimates. Enterprise clients use Perplexity Search for deep financial research and insights at scale. These audiences have previously been served by completely different infrastructure.
"Perplexity Finance has been built to break down those silos, giving both groups access to high-quality, real-time data, and the ability to research deeply using the same premium data sources that are usually only accessible to institutional investors."
To serve both audiences, Perplexity needed infrastructure that could deliver institutional data quality with API-native simplicity.
Trust through first-party information and global coverage
Perplexity's mission is "to serve the world's curiosity, which is what we've built into our experience for every investor."
In practice, that mission translates to a specific user experience requirement: users shouldn't need to leave Perplexity to verify or deepen their research.
"You're not just pulling statistics and then switching tabs and context to find the answers that are relevant to you; it can all be done in one place, with Perplexity providing support to explore even further than traditional search experiences are built for."
This creates a fundamental data challenge for AI-powered financial research platforms. When a user asks about a company's earnings results, management guidance, or strategic direction, the output must be grounded in verifiable, first-party information. Not web summaries, third-party interpretations, or delayed feeds.
Finance is a domain where it's not good enough to be 99% or even 99.9% accurate – users need to be able to trust your data fully.— Jeff Grimes, Head of Live Events at Perplexity Finance
Earnings calls represent the highest-stakes moment for this requirement. Management speaks publicly about results, outlook, and strategy. Analysts ask probing questions. The accuracy matters. The timing matters. The completeness matters.
To make earnings research viable inside an AI product, Perplexity needed three things:
Live delivery – Real-time transcripts and audio as events happen
First-party sources – Direct from companies, verbatim, with no intermediary interpretation
Global coverage – Consistent experience across markets, company sizes, and event types
Quartr delivers all three.
Why Quartr: Comprehensive data, modern architecture
"Quartr gives us the ability to deliver a comprehensive earnings experience to our users – live transcripts, live audio, and company-provided documents."
More specifically, Perplexity integrated six Quartr datasets:
Live audio: Streaming access to events as they happen, enabling users to listen directly alongside AI-generated summaries.
Live transcripts: Real-time transcription delivered with speaker attribution and timestamps, allowing citation to specific moments and quotes.
Backlog audio: Extensive archive of high-quality audio from earnings calls and unique events like capital markets days, conferences, annual general meetings, and M&A announcements.
Backlog transcripts: High-accuracy transcripts from earnings calls and unique events.
Filings and reports: Annual and quarterly earnings reports, 10-Ks and 10-Qs, and unique events.
Slide presentations: High-resolution PDFs of slide presentations from earnings calls and unique events. Scannable for any keyword or data point.
Instead of integrating multiple solutions – one vendor for transcripts, another for audio, a third for filings – Quartr provides a unified API covering the full earnings experience.
"We use Quartr for live earnings transcripts, live earnings call audio, and company documents. Paired with Perplexity's expertise in retrieving fresh web content and synthesizing into powerful insights, these three Quartr features make Perplexity a go-to destination for following earnings calls."
API-first delivery meant Perplexity's engineers could test endpoints, validate data structures, and build integration logic using familiar tools and workflows.
"Straightforward, easy collaboration in Slack channel, straightforward integration via REST API endpoints. The Quartr team is always responsive to questions. The API was very easy to integrate, too."
"Finance is a domain where 99.9% accurate isn't good enough"
The difference between a few percentage points in accuracy seems marginal until you operate at scale during earnings season. A single misattributed quote creates liability. A transcript delay of five minutes means analysts miss the guidance revision that moves the stock.
"Quartr's quality control process helps us maintain accuracy and trustworthiness."
This level of accuracy enables AI-generated earnings insights to be used with far less manual verification in professional workflows.
"Users regularly tell us that the earnings feature of Perplexity Finance is one of their favorites."
What this means for enterprise buyers
For teams evaluating financial data infrastructure, the Perplexity case offers several decision-relevant signals:
First-party data: By grounding AI outputs in verified first-party transcripts and documents, Perplexity developed features trusted by both professional and retail users.
Ease of use: By covering live transcripts, audio, and documents through one API, Quartr eliminated the coverage inconsistencies and integration complexity that typically come with multi-vendor strategies.
Operational reliability: Reduced dependency on multiple or backup vendors.
For AI products in finance, these factors compound. Data quality determines output quality, and vendor reliability determines user trust.
Looking ahead: Building on a proven foundation
Perplexity's roadmap reflects continued investment in financial research depth.
"As we continue toward our mission of helping users do their best financial research, we will be integrating more tools directly into Perplexity answers, making them even more comprehensive."
The team maintains a deliberate near-term focus.
"We often say internally that we don't think in 5-year plans, or even 1-year plans." Instead, Perplexity optimizes for continuous expansion of capability and user value."
"Our right now plan is to continue to expand the Perplexity Finance experience in a way that we continue to deliver an experience that's second to none to every type of investor: casual, retail, institutional, and beyond."
With Quartr as the core earnings data layer, that roadmap is anchored in first-party information delivered live and at global scale. Future enhancements will build on this foundation, including deeper integrations with financial documents, expanded event coverage, and more sophisticated AI tools for analyzing management communications over time.
For enterprise teams building financial research products, the lesson is clear: your AI is only as good as your data, and your data strategy is only as good as your infrastructure partners.
Perplexity chose to build on first-party sources, comprehensive earnings coverage, and production-grade reliability. The result is a platform where users across every segment, from retail investors to institutional researchers, choose to do their earnings research.