Overview

Storytell can enhance your internal data with almost any type of external web data. Instead of generic web search, you get category-intelligent search: specialized retrieval that understands the structure of different content types. Each category uses optimized logic for how that content is indexed and consumed, so you can run multi-source analysis workflows that combine your private knowledge with precisely targeted external intelligence.

How to turn on public data sources

You control whether a chat can use public data (web search) in the Powering This Chat panel. When public data is on, Storytell can search the web—people, companies, news, research, GitHub, SEC filings, and more—to answer your prompts.

To turn on public data sources:

1. Open the chat you want to use public data in.

2. Open the Powering This Chat panel (the side panel that shows which knowledge and tools power this chat). You can open it from the chat header or the side-panel toggle.

3. In the panel, find the Use Public Data In This Chat card. It lists the public source types (Email Finder, Web, People, Companies, Research Papers, Social Media, YouTube, Podcasts, News, GitHub, SEC Filings).

4. Turn the switch on. When it’s on, the label shows Use Public Data In This Chat and the card is highlighted; when it’s off, it shows No Public Data Will Be Used In This Chat.

Your choice is saved for that browser. You can turn public data off anytime by toggling the same switch off. If you add collections or files to the chat, Storytell will use both your knowledge and public data (when enabled) to answer.

What are public data sources?

Public data sources are external, web-accessible information that Storytell can search and cite when answering your prompts. When you don’t scope a chat to specific collections or files, or when you explicitly allow public knowledge, Storytell may call on these sources to ground answers in real-time, factual data from the web.

You can use public data sources to:

• Validate internal assumptions against market signals and news

• Discover people, companies, research, and code relevant to your work

• Monitor sentiment, trends, and regulatory or financial disclosures

• Combine multiple categories in one analysis (e.g., companies + news + SEC filings)

Public data sources work alongside your knowledge base (collections and files). Storytell prioritizes your scoped assets and layers in public data when it improves the answer.

Public data source categories

Email Finder

What it finds: Professional email addresses associated with people or organizations.

Optimized for: Contact discovery, outreach lists, lead enrichment.

Example prompts:

• "Find the email address of Storytell.ai’s CEO”

Why it matters: When building prospect lists or validating contacts, Storytell can surface professional emails and cross-reference them with your internal CRM or lead data.

Web

What it finds: General web pages, articles, and sites—no single content type.

Optimized for: Broad discovery, how-to content, comparisons, and anything that doesn’t clearly fit another category.

Example prompts:

• "Best practices for remote team retrospectives"

• "Comparison of project management tools for engineering teams"

• "Recent blog posts about developer experience"

Why it matters: Use Web when your intent is general. For more precise results, choose a specific category (e.g., News, Research Papers, Personal sites) when it fits.

People

What it finds: Individuals by role, title, company, or expertise—professional profiles and organizational context.

Optimized for: Expert discovery, org charts, competitive intelligence, hiring research.

Example prompts:

• "AI researchers at Google working on transformer architectures"

• "Executives at companies building autonomous vehicles"

• "Engineers working on Rust at major tech companies"

Why it matters: When analyzing competitive landscapes or sourcing expertise, Storytell can identify key people, check credentials against your internal research, and map professional networks in one workflow.

Companies

What it finds: Businesses, startups, and organizations—corporate info, products, and market positioning.

Optimized for: Competitive analysis, market mapping, vendor and partner discovery.

Example prompts:

• "Startups building AI code assistants"

• "Companies in the autonomous vehicle space with recent funding"

• "B2B SaaS companies targeting healthcare compliance"

Why it matters: Combine this with your strategy docs and market research. Storytell can pull live competitive intelligence from the web and cross-reference it with your internal assumptions—validating market sizing, finding new competitors, or surfacing acquisition targets.

Research Papers

What it finds: Academic and peer-reviewed literature—studies, methodologies, and citation networks.

Optimized for: Literature review, methodology validation, R&D and grant writing.

Example prompts:

• "Recent advances in retrieval augmented generation"

• "Peer-reviewed studies on transformer architectures for NLP"

• "Clinical trials for cancer immunotherapy published in 2025"

Why it matters: When writing whitepapers, proposals, or technical docs, Storytell can surface the latest research, extract methodologies, and validate claims against the academic literature alongside your internal R&D notes or product specs.

Social Media

What it finds: Posts, threads, and reactions from social platforms—public sentiment and real-time discussion.

Optimized for: Sentiment, viral trends, community and customer reactions.

Special capability: Higher result limits (e.g., up to 50 results where supported) to capture more short-form posts.

Example prompts:

• "Developer reactions to new React features"

• "What are people saying about GPT-5?"

• "Public sentiment about recent product launch"

Why it matters: Use this as an early-warning and feedback channel. Storytell can monitor sentiment around competitors, track reactions to your announcements, and compare social signals with your internal metrics or customer feedback.

YouTube

What it finds: Videos, channels, and video metadata—tutorials, talks, demos, and commentary.

Optimized for: How-to content, conference talks, product demos, and expert commentary.

Example prompts:

• "Conference talks on microservices architecture 2024"

• "Tutorials on building RAG applications"

• "Product demos for AI coding assistants"

Why it matters: When you need visual or narrative context—demos, talks, or step-by-step guides—Storytell can surface relevant YouTube content and tie it to your internal docs or training materials.

Podcasts

What it finds: Podcast episodes and show metadata—interviews, discussions, and long-form audio content.

Optimized for: Expert interviews, industry trends, and narrative or opinion content.

Example prompts:

• "Podcast episodes on startup fundraising in 2024"

• "Interviews with CTOs about platform engineering"

• "Discussions about AI regulation and compliance"

Why it matters: Podcasts often capture nuanced, expert perspectives. Storytell can find relevant episodes and summarize or cite them in the same workflow as your internal notes or strategy docs.

News

What it finds: Current events, breaking stories, press releases, and journalism.

Optimized for: Timeliness, editorial content, and news cycles.

Example prompts:

• "Latest AI regulation developments in the EU"

• "Recent tech layoffs at major companies"

• "Breaking news about quantum computing breakthroughs"

Why it matters: Layer live news over your planning documents. Ask Storytell how recent regulatory changes affect your roadmap or how competitor announcements align with your strategic assumptions.

GitHub

What it finds: Repositories, libraries, and open source projects—code, READMEs, and community activity.

Optimized for: Technical evaluation, implementation patterns, and ecosystem research.

Example prompts:

• "Best Go testing frameworks with recent activity"

• "Python ML libraries for time series forecasting"

• "React component libraries with TypeScript support"

Why it matters: When evaluating technical choices or doing vendor research, Storytell can survey the open source landscape, assess community momentum, and compare findings with your internal architecture or tech stack decisions.

SEC Filings

What it finds: SEC filings, 10-K/10-Q reports, earnings materials, and other regulatory disclosures.

Optimized for: Financial analysis, risk factors, and investor-related content.

Example prompts:

• "Apple Q4 2025 earnings report"

• "Tesla SEC filings mentioning battery technology"

• "10-K risk factors for autonomous vehicle companies"

Why it matters: When building financial models or investment theses, Storytell can pull audited data from filings, extract risk factors, track M&A activity, and validate assumptions against what public companies report—while still reasoning over your internal projections.

How search works

Public data search is semantic: it interprets meaning and intent, not just keywords. The system handles synonyms and related phrases automatically.

Example:
The query "What are the latest developments in quantum computing?" can surface results about quantum breakthroughs, qubit advances, algorithm progress, and innovations—without you listing every possible phrase.

The semantic layer understands that:

• "Machine learning" can match "ML," "artificial intelligence," and "neural networks"

• "Affordable flights" can surface "cheap airfare" and "budget travel"

• "Engineers at Amazon" targets employee-style profiles, not generic job postings

FAQ

Will my chat still use my collections and files when public data is enabled?
Yes. When you scope a chat to collections or files, Storytell prioritizes that knowledge and uses public data to augment or validate. If you don’t specify scope, it may rely on public knowledge and general training data.

What happens if I don’t select a category?
Storytell can still search the web using a general (Web) or auto-selected strategy. For clearer intent and better results, choose a category when it fits your question (people, companies, news, etc.).

Can I use multiple categories in one prompt?
Yes. You can ask for an analysis that pulls from companies, news, SEC filings, and social media in a single answer. Storytell will orchestrate the right searches and synthesize the results.

Why does Social Media return more results than other categories?
Social posts are short. Allowing more results (e.g., up to 50) makes it easier to capture meaningful sentiment and variety; other categories typically use a smaller default (e.g., 10) for longer documents.

Are public data sources available on all plans?
Availability may depend on your plan. Check your workspace settings or contact your admin if you don’t see web search or category options.

:::important
Anyone with access to a shared chat or workspace can see that public data sources may be used when scope is not restricted. Be careful about sharing links if you want to limit what external data is included in answers.
:::