CivicScience’s new consumer insights product addresses generative AI trust

CivicScience, an industry leader in providing real-time consumer insights, has introduced a generative AI analytics assistant that establishes a secure environment for making crucial business decisions by resolving the unique trust concerns associated with this rapidly expanding technology.

CivicScience founder and CEO John Dick mentioned that GenAI is transforming how corporate leaders derive value from data. 

Least reliable in situations where precision is of the utmost importance are large language models (LLMs), which frequently experience hallucinations and utter credible fiction (e.g., OpenAI’s ChatGPT and Google’s Bard).

To mitigate these issues, Sage implements a “AI sandwich.” By deciphering user inquiries and summarising findings, ChatGPT functions as the bread. Consisting of five billion responses from millions of American consumers and nearly 600,000 poll questions entered into a database, CivicScience’s proprietary systems conduct real-time statistical analyses.

Sage users employ clear and concise language to inquire about consumer analytics, for instance, “What are the prevailing trends in the Northeast among middle-aged women when it comes to purchasing luxury automobiles?” In addition to a spreadsheet download containing expanded underlying data, they receive responses with abundant citations that direct readers to the source data in CivicScience’s InsightStore. Additionally, each AI-generated response is assessed and rated by the system in order to verify its accuracy.

Dick continued, “GenAI will ultimately free data-driven organisations from the tedium of data processing, allowing them to concentrate more on the ‘why’ behind insights.” “Sage eliminates a substantial amount of concern by verifying and associating each number with data that has been relied upon by our clients for years.” Sage also resolves the issue of how CivicScience can capitalise on its daily dialogues with 500,000 American consumers, which generate over four million fresh poll responses on a daily basis.

“For the majority of businesses, GenAI is merely an enhanced search engine that adeptly summarises inert documents and establishes relationships between various texts,” said Joseph Galarneau, Chief Product Officer at CivicScience. “However, this methodology proves inadequate when confronted with the exponential growth of enormous data streams daily.” “Our research efforts are concentrated at the intersection of LLMs and dynamic data, where we believe there is enormous potential.”

Sage, which was just two weeks ago introduced as one of the initial martech products to utilise OpenAI’s next-generation LLM, is presently accessible through a webchat interface and will soon feature integrations with Slack and Microsoft Teams, enabling enterprise employees to communicate with it. The system is capable of comprehending inquiries in over a hundred different languages and will eventually possess the capability to engage in conversation in a significant portion of them. Sage has been in beta with over thirty companies, including several Global 2000 firms, since the summer.

Sage represents the most recent artificial intelligence (AI) offering from CivicScience, an organisation that was established in an incubator at Carnegie Mellon University, a leader in global computer science, over a decade ago. The organization’s QGen poll generation platform, which integrates GenAI with human review for each question, exhibited a tenfold enhancement in internal process efficiency and AI safety measures earlier this year. Years ago, the organisation implemented supervised and unsupervised machine learning in its InsightStore analytics application.