THE integration of large language models (LLMs) in chatbot development is on the rise, as reported by data cloud company Snowflake. With an increasing number of businesses tapping into this technology, chatbots now constitute 46% of LLM applications, a significant jump from the previous 18% recorded in May 2023.
Moreover, insights gleaned from a survey conducted within the developer community of Streamlit, an open-source hub for machine learning data science web apps, revealed that approximately 65% of respondents are utilising LLM projects for professional purposes. This indicates a notable shift towards leveraging generative AI to enhance workforce productivity, efficiency, and analytical capabilities.
These findings, drawn from usage data spanning over 9,000 Snowflake customers, are detailed in Snowflake’s latest Data Trends 2024 report. The report highlights how global enterprise leaders are harnessing resources like AI to fortify their data infrastructure and revolutionise future business operations. Notably, there has been a transition from text-based input LLM applications (82% in 2023) to chatbots with iterative text input, facilitating more natural conversations (54% in 2024).
Jennifer Belissent, Principal Data Strategist at Snowflake, explains the surge in conversational apps, attributing it to the innate human inclination towards interactive communication. Belissent anticipates this trend to persist, especially with the growing ease of building and deploying conversational LLM applications, backed by robust data governance and security measures. She anticipates that these dynamic and adaptable chatbots will cater to both business requirements and user expectations.
The report further reveals that within the past nine months, 20,076 developers from Snowflake’s Streamlit community have collectively developed over 33,143 LLM apps. Python remains the preferred programming language for AI projects due to its user-friendly nature, extensive developer community, and vast library ecosystem.
In terms of development platforms, there’s a notable trend towards building LLM applications directly on the data management platform. Snowflake witnessed a remarkable 311% increase in native app development between July 2023 and January 2023. This streamlined approach not only expedites application development and deployment but also minimises operational overheads by eliminating the need for data transfers to third-party technologies.
