Recent advancements in AI have democratised creative endeavours by providing tools that assist individuals in writing, coding, and even generating images, by drastically reducing barriers to entry. This technology has already been applied to data analysis, but such tools primarily focused on turning natural language into Structured Query Language (SQL) prompts. As a result, they are unusable for companies using other kinds of data storage, including NoSQL databases, corporate CRMs, and even third-party APIs.
Germany-headquartered PandasAI makes all of these sources accessible, even to those unfamiliar with SQL or professional data analysis tools. Now, the company building a conversational AI Assistant for data analysis has raised $1.1 million in pre-seed funding from Runa Capital, Episode1 (which backed Phasecraft and Creator Fund), and Vento. PandasAI will use the proceeds of the round to aggressively ramp up its hiring and talent acquisition.
“PandasAI’s vision is to render data analysis accessible, enabling decision-makers to swiftly and effectively make data-driven decisions,” said Gabriele Venturi, Founder of PandasAI. “We are building an AI assistant that won’t only use data to use answer questions, but will also help users to clarify their original ask, when needed, to ask the right questions. Companies will be able to deploy it in their private environments and generate insights related to their specific organisation.”
“Gabriele’s execution and the support of the PandasAI community have established it as the leading open-source solution for conversational data analysis,” said Maxime Corbani, a Senior Associate at Runa Capital based in Paris. “Data analysts still spend a lot of time gathering, cleaning, merging and analyzing internal data. PandasAI automates the entire pipeline and helps produce insights.”
Builds AI assistant for data analysis
Founded by Gabriele Venturi, PandasAI enhances the widely-used Python data analysis library pandas by introducing a conversational aspect. Instead of complex queries, users can ask questions using natural language to generate graphs and extract powerful insights from public data sets or companies’ data lakes.
PandasAI allows users to access well-known Large Language Models (LLM), like GPT-4, on top of enterprise data without the need to send the data to OpenAI or other generative AI platforms.
PandasAI is exploring integration possibilities with other language models such as LLama 2, and is developing BambooLLM, a LLM focused on data analysis, as a complementary solution. It is an open-source project under the MIT license with more than 8.5K stars on GitHub.
To get started with PandasAI, it just needs to be installed. Users can then begin to effortlessly integrate it into their Python coders. By combining Generative AI with Python’s pandas library, it allows users to query and analyse data using natural language, facilitating tasks such as graph generation and cross-dataframe correlations.
