AI Integration Revolutionized: New Framework Simplifies Software Development with a single Line of Code
Developers can now seamlessly incorporate the power of large language models into their applications without the complexities of manual prompt engineering, thanks to a groundbreaking open-source framework called byLLM.
The era of tedious, complex prompt engineering might potentially be coming to an end. A new open-source framework, byLLM, promises to dramatically simplify the integration of large language models (LLMs) into existing software, requiring as little as a single line of code. The innovation, detailed in a conference paper presented at the SPLASH conference in Singapore in October 2025 and published in the Proceedings of the ACM on Programming Languages, automatically generates context-aware prompts, eliminating the need for developers to painstakingly craft them by hand.
“This work was motivated by watching developers spend an enormous amount of time and effort trying to integrate AI models into applications,” explained Jason Mars, an associate professor of computer science and engineering at the University of michigan (U-M) and co-corresponding author of the study.
The challenge of integrating AI lies in the essential differences between conventional programming and LLM operation. Traditional programming relies on explicitly defined variables, while LLMs process natural language text. Bridging this gap traditionally required developers to act as “translators,” manually constructing textual input – a process known as prompt engineering – which can be both time-consuming and imprecise.
byLLM, along with its supporting runtime, automates this crucial step. “This innovation makes integrating powerful AI models into software as easy as calling a function, so developers can focus on building creative solutions rather than wrestling with prompt engineering,” stated Lingjia Tang, an associate professor of computer science and engineering at U-M and co-corresponding author of the study.
How byLLM Bridges the Gap Between Programming and AI
At the heart of byLLM is the “by” operator, which acts as a bridge between conventional code and LLM operations. A compiler,leveraging a meaning-typed intermediate representation,gathers semantic information about the program and the programmer’s intentions. This information is then converted into a focused set of prompts by an automatic runtime engine,directing the LLM’s processing.
Evaluations demonstrate that byLLM surpasses existing prompt engineering frameworks, such as DSPy, in terms of accuracy, runtime performance, and robustness. A user study revealed that developers using byLLM completed tasks more than three times faster and wrote 45% fewer lines of code.
“By inferring the programmer’s intent from code snippets, byLLM lowers the barrier for AI-enhanced programming and could enable an entirely new wave of accessible, AI-driven applications,” said Krisztian Flautner, a professor of engineering practice at U-M and co-corresponding author of the study.
The team has made byLLM open source, and it has already garnered meaningful traction, with over 14,000 downloads within a single month and growing interest from industry partners. This accessibility has the potential to empower smaller teams and even non-expert programmers to create refined AI applications, fostering innovation in areas like personalized software, research tools, and interactive learning environments.
Companies across diverse sectors – including finance, customer support, healthcare, and education – can leverage this approach to rapidly integrate LLMs into their products, significantly reducing engineering overhead.
More information about byLLM can be found at: https://github.com/jaseci-labs/jaseci/tree/main/jac-byllm. The research paper, “MTP: A Meaning-Typed Language Abstraction for AI-Integrated Programming,” is available in the Proceedings of the ACM on Programming Languages (DOI: 10.1145/3763092) and on arXiv (DOI: 10.48550/arxiv.2405.08965).
This breakthrough promises to democratize AI integration, ushering in a new era of accessible and efficient software development.
