Who is Narrative Science?

We are a technology company based in Chicago, with offices in New York and D.C. What began as an artificial intelligence research project at Northwestern University is now a rapidly growing SaaS business, having raised over $32 million from leading investors Battery Ventures, Sapphire Ventures, In-Q-Tel, Jump Capital and USAA. Our patented advanced natural language generation (NLG) platform, Quill, is helping companies across multiple industries transform data into narratives that improve decision-making, increase employee productivity and provide a better customer experience - all through the power of storytellingLearn more ›

What is Quill?

Quill is our patented advanced NLG platform, fueled by artificial intelligence, that automatically creates data-driven communications at machine scale. It analyzes data from disparate sources, understands what is important, then generates perfectly written narratives to convey meaning for any intended consumer or business audience. Wherever companies are spending too much time on data-driven report writing workflows, Quill automates these processes and eliminates manual involvement. Quill also creates new revenue opportunities by transforming data into engaging content that can be productized and monetized. Learn more ›

Is Quill artificial intelligence?

Quill is powered by artificial intelligence, enabling it to write narratives indistinguishable from what a human would write at unprecedented scale. Our patented technology is a unique synthesis of data analytics, reasoning and narrative generation. Learn more ›

What is Natural Language Generation (NLG)?

Natural Language Generation (NLG) is a subfield of artificial intelligence (AI) which produces language as output on the basis of data input. Most NLG systems are basic in that they simply turn data into text, whereas our advanced NLG platform transforms data into narratives - personally relevant stories that have a logical beginning, middle and end - of any length. Our patented methodology starts by understanding your communication goals and the metrics necessary to meet those goals, performs the relevant analysis to determine what is most interesting and important, and identifies the required data needed to generate the narrative to your intended audience.

How is Natural Language Processing (NLP) different than Natural Language Generation (NLG)?

Fundamentally, NLP reads while NLG writes. NLP systems look at language and figure out what ideas are being communicated. NLG systems start with a set of ideas and turns them into into language that, in turn, communicate them. Advanced NLG systems start with raw structured data and figure out those ideas before turning them into narratives.

Why should I work with Narrative Science?

We are the leading provider of advanced NLG solutions for the enterprise. Our patented technology platform crafts the voice of data and analysis into powerful communication that is brand-aware and compliant. We hire talented employees with domain expertise who are focused on developing in-demand solutions and providing exceptional experiences. We strive to get our customers up and running fast so they can quickly start realizing the benefits of our software and services. Contact us to discuss the best place to start ›

Who are Narrative Science’s clients?

We are a rapidly growing business with over 70 enterprise customers including USAA, MasterCard, Deloitte, Credit Suisse, American Century Investments and members of the U.S. Intelligence Community. Over 12,000 small business users are already utilizing our self-service applications. Learn more ›

Is Narrative Science hiring?

Yes! Our greatest asset is our people. If you are interested in pursuing a career in a flexible, energetic and inspiring environment, we invite you to learn more about our opportunities in sales, marketing, product management, professional services and engineering. Explore opportunities ›

Data and Analysis

What types of data does Quill ingest?

Quill ingests any type of structured data in any of the most prevalent formats (JSON, XML, CSV, etc.).

Can Quill leverage multiple data sources?

Yes. Quill can consume data from disparate data tables and databases if the desired narrative is reliant upon these sources. Upon consumption, Quill will aggregate the data, perform various analytics and calculations in order to produce a more concise dataset, and use this dataset to draw deeper insight and ultimately produce the appropriate narrative. Quill can also take the output from business intelligence and analytics platforms and generate a narrative to explain the conclusions drawn from those platforms’ analyses. Quill’s intelligent narratives can accompany visualizations and summarize disparate reports, making it easy for any employee to quickly understand business performance and take action.

How does Quill obtain data?

Quill can ingest data in a multitude of ways. Customers usually choose to deliver data via our web services API for on-demand processing, or or to our SFTP server for batch processing. For our on-premise customers, Quill can retrieve and ingest data directly within a customer owned data center or a customer managed cloud environment. Learn more ›

How and where is data processed?

For those customers who utilize our SaaS offering, customer data is processed in our database hosted in the Amazon Web Services (AWS) cloud, an extremely reliable and secure cloud computing environment that enables customers to deliver data quickly and securely. Data is wiped from the database after a certain time window in order to support Professional Services iteration with customers during development. Customers that utilize our on-premise offering have full control of their data and the data is processed within their own managed environment. Learn more ›

Is the data secure?

Yes. We take data security very seriously and uses industry standard encryption protocols to assure that data travelling to and from its hosted infrastructure will not be compromised. We have achieved certifications in SOC 2 compliance and conduct an annual third party audit to validate the security of our infrastructure. Access to Narrative Science servers and the Quill application is restricted to valid personnel through firm access control protocols and firewalls. We also offer on-premise deployment options for those customers that want full control over security features or are hesitant to hand over data to a third party service provide.

What is Narrative Analytics?

Narrative Analytics is our patented approach to analysis that is completely driven by your communication goals. By understanding the purpose of your communication, identifying the metrics to meet that purpose, performing the analysis to figure out what is most interesting and important, and pulling in the relevant data, Quill generates a narrative to your intended audience that is indistinguishable from what a human would write. Learn more ›


What are the different types of narratives Quill can produce?

Quill can produce content about virtually any topic, as long as that content is rooted in quantifiable data. The platform can produce diverse written assets directly from data—including BI reports, regulatory reporting, web content, Tweets, customer communications and much more. Narratives are adaptable to the desired length, tone, and style, depending on the intended audience. Learn More ›

Is Quill template-driven?

No. Unlike basic NLG systems, which simply turn data into text and populate into pre-defined templates, Quill is an Advanced NLG platform that generates narratives driven by the purpose of a particular communication.

Can Quill match my tone, voice, and style?

Yes. Quill can be trained to produce content of a particular voice and tone (i.e. colloquial or formal) upon customer request. Quill can also be trained to mimic the voice and tone of reports and communications you are already producing. The style is completely dependent on your needs, with narrative output including anything from bullet points, long-form summaries, emails, tweets, web content, etc.

How fast is Quill?

The rate at which Quill produces content varies depending on the size and complexity of the data, however in most cases narratives are produced in seconds.

How and where does Quill deliver narratives?

Narratives can be generated in a variety of formats, including Word, PDF, HTML, JSON, XML, or plain text, and can be delivered via the return of a customer initiated API call for on-demand delivery, or picked up from a Narrative Science SFTP server for batch delivery. Quill’s narratives can be published into websites, mobile, social, business intelligence platforms, and any application that can connect to our API.

Can Quill produce visualizations?

Yes. Quill can produce charts and graphs to accompany your narrative when requested, or the narratives can be deployed alongside your existing charts and graphs to provide further explanation of their insights.


Is Quill offered as a Software-as-a-Service (SaaS)?

Yes. Narrative Science hosts Quill in Amazon’s Platform as a Service offering, Amazon Web Services, allowing us to quickly scale our infrastructure in response to varying customer story requests. Customers interact with the hosted platform when sending data, retrieving stories, and reviewing content in order to validate quality or request changes. Quill itself its configured by a team of Solutions Consultants who are well versed in both the customer communication goals and the process of training Quill to produce the intended content. Learn more ›

Is Quill offered as a deployment option on our premises?

Yes. For those customers that want full control over security features, or are hesitant to hand over data to a third party service provider, the full stack of Quill services can be installed within a customer managed environment. This allows narrative generation to take place without data ever being exposed to a public network. Quill can be deployed into either a customer owned data center or a customer managed cloud environment. Learn more ›

Once we’ve decided to work with Narrative Science, how long does it take to start running?

Timelines can vary anywhere from 2 to 12 weeks. It depends upon the length and complexity of the desired narrative, the size and complexity of the data, and whether Quill is familiar with the domain. Other factors to consider are the available resources on the client side necessary to meet the desired data specification and review content in order to validate quality. If the client does not have those resources, we are happy to provide assistance through our professional services team.

How do I get started with Narrative Science?

We would love to talk with you and show you the power of Quill. Please contact us, request a demo, or check out our various solutions to learn how Quill can be applied to increase employee productivity, improve customer engagement, drive innovation and ensure compliance. Learn more ›

Where can I see Quill in action?

To see Quill in action, you can request a demo, check out our resource page to view Quill examples or try our free applications Quill Engage (if you have a Google Analytics account) and Quill Connect (if you have a Twitter account).

Our Quill Engage application transforms your Google Analytics (GA) data into insightful, easy to read narratives about your website performance, with helpful tips on how to improve; simply authorize your GA account to receive your narrative. Our Quill Connect application transforms your Twitter data into a story about your Twitter activity with insights about your followers and targeted recommendations on how to increase your sphere of influence; simply authorize your Twitter account to receive your narrative.