Tellius, an AI-powered decision intelligence platform has announced Copilot, a new set of GPT-integrated product enhancements that leverage OpenAI’s large language models. The new features automatically generate natural language synonyms, code improvements, and enhanced data stories, adding to the company’s existing AI-powered analytics capabilities to further enable every insights-driven organisation to implement and scale self-service analytics.
Tellius Copilot will power the following new experiences in the platform:
- Recommended data dictionaries. With a single click, Copilot generates synonyms, descriptions, and business-friendly names for data columns. This feature enhances the natural language search capability to deliver contextual insights to anyone in the organisation.
- Code generation, improvement, and validation. When utilising SQL/Python code to transform data and apply advanced business rules, Copilot automatically identifies errors and suggests code fixes, adds comments and explanations to code, and even generates code from scratch from natural language input. This new feature makes it easier and faster for data experts to prepare data for analysis.
- Enhanced data storytelling. In addition to the product’s AI-generated insights and best-fit visualisations, Copilot goes one step further to provide narratives that summarise key insights that are easy for users to understand. Business teams can easily interpret data analysis and make decisions with confidence.
“Applying AI to data analytics has always been in our DNA, and integrating generative AI and LLMs is another step in eliminating friction in the analytics workflow and enabling customers to take their self-service analytics and data literacy initiatives to the next level,” says Ajay Khanna, founder and CEO of Tellius. “Copilot is a natural extension of the Tellius platform to further democratise data access and speed up the journey from data to decisions.”
While the applications of GPT has the world abuzz with its impressive ability to generation human-like text responses from a broad array of prompts, it has limitations in analytics use cases due to its a) inability to handle complex or multiple data sources; b) lack of data visualisations; c) inability to answer questions about an organisation’s data and business without additional fine-tuning; d) and documented tendency to return incorrect answers, all of which could lead to wrong conclusions, bad decisions, and reduced trust from end users.
“Tellius addresses inherent limitations and ensures that users can trust the answers they receive by combining GPT’s strengths of context and summary generation with our own real-world tested AI algorithms and proprietary logic. Providing accurate and reliable data insights to our users is a core tenant we would never waiver from,” says Hardik Chheda, chief product officer at Tellius. “Additionally, while GPT and other LLMs are extremely promising new technologies, Tellius takes a human-in-the-loop approach to ensure that automation never takes away the control and governance that our users have come to love with our platform.”
Tellius Copilot is currently in limited preview as an opt-in product enhancement.
To learn more, register for the product webinar on April 19, 2023, or sign-up for early access. To get a preview at the Gartner Data & Analytics Summit in Orlando, FL, March 20-22, 2023, schedule a meeting here.
Follow us and Comment on Twitter @TheEE_io