大语言模型/基础模型的快速发展正在对软件和软件工程产生着深刻而全面的影响。ai赋能的软件（aiware）正在为软件工程带来新的机遇和挑战。本论坛邀请来自全球范围学术界和工业界的多位专家和学者，共同探讨在大语言模型/基础模型时代下软件和软件工程的未来。本论坛重点关注ai赋能的软件应用和面向ai赋能软件的软件工程，包括ai赋能软件的构建开发、维护与平台支撑等，以及产业实践中所面临的挑战，并探讨该方向未来的发展机遇。欢迎软件工程、人工智能等领域的专家和学者以及所有感兴趣的同行们前来参加。注：本论坛所有报告现场均提供中英文实时翻译。
01
论坛组织委员会
娄一翎 （复旦大学）
thomas zimmermann （microsoft research）
谢涛 （北京大学）
ahmed e. hassan（queen’s university）
02
论坛议程
03
论坛报告嘉宾简介
ahmed e. hassan
报告题目：aiware: re-thinking software and software engineering in the foundation model/llm era
报告摘要：
“software for all by all” is the future of humanity. aiware, i.e., ai-powered software, is democratizing software creation. we must reimagine software and software engineering (se), enabling individuals of all backgrounds to participate in its creation with higher reliability and quality while leveraging foundation models (fms), such as large language models (llms).
the unique properties of fm-powered applications (fmware), like prompts and agents, coupled with the intrinsic limitations of fms (e.g., hallucination) bring a completely new set of software engineering challenges. in this talk, i will discuss several key challenges that have caused enterprise fmware development to be unproductive, costly, and risky. i will also discuss the road ahead for applying fmware innovations in the context of the se domain.
报告人简介：
ahmed e. hassan is an acm fellow, ieee fellow, and an nserc steacie fellow. he is a laureate of the mustafa prize, commonly likened to a nobel, for founding the field of ai-augmented se/intelligent se and its worldwide impact on society and industry. he is the only individual globally to be recognized as a distinguished/influential educator by both of the world’s leading professional se societies, ieee-tcse and acm-sigsoft. he is a canada research chair (crc) in software analytics, and previously a nserc/blackberry software engineering chair. today he leads worldwide efforts to rethink software and se in the llm/fm era. see https://www.aiwarebootcamp.io, https://fmse.io, https://www.aiwareconf.org, and https://arxiv.org/abs/2402.15943 for more details. early systems and technologies developed by hassan and his teams are already integrated into products used by millions of users worldwide. his industrial experience includes architecting and scaling the global blackberry wireless platform, as well as working for ibm research at the prestigious almaden lab in usa, and the computer research lab of nortel networks. dr. hassan is the named inventor of patents in several jurisdictions around the world including the united states, europe, india, canada, and japan.
谢涛
报告题目：platform-software support for the development and deployment of large-language-model-powered software applications
报告摘要：
in recent years, large language models (llms) have attracted extensive attention in the research community and industry, and have gradually shown their great potential in various fields. as a new form of software applications, the development and deployment of llm-powered software applications pose new challenges to software engineering and system software. this talk presents some of our recent research efforts on platform-software support for development and deployment of llm-powered software applications, including a version control system for agile collaborative development of llms, and the runtime system design of llm-based agents. the talk also discusses the relationship between llm and traditional software, llm engineering in the age of “software engineering 2.0”, and opportunities for deeply integrating traditional software and llm with platform-software support.
报告人简介：
tao xie is a ccf fellow, director of ccf technical committee of system software (tcss), peking university chair professor, chair of the department of software science and engineering in the school of computer science at peking university, chief scientist of beijing institute of open source chip, deputy director of the key lab of high confidence software technologies, ministry of education. he was a full professor at the department of computer science, the university of illinois at urbana-champaign (uiuc), usa. he is a foreign member of academia europaea, and a fellow of acm, ieee, and aaas. he serves as the risc-v international ai/ml sig chair, risc-v ai computility ecosystem (race) committee chair, and risc-v summit china 2024 conference co-chair.
dayi lin
报告题目：fmarts: a lifecycle engineering and productivity platform for foundation model/llm powered software (fmware)
报告摘要：
foundation models (fms) such as llms have led to a new generation of software (fmware). despite the hype around their applications, the industry still faces tremendous challenges to productively develop and deploy high quality and trustworthy fmware at scale. such challenges require a systematic rethinking and full-stack technology innovation of the new generation of software and its engineering paradigm. in this talk, we provide an overview of our work on fmarts, a full-stack fmware engineering platform that synergizes lifecycle environment (i.e., development, version management, semantic observaibilty, debuggability, and optimization of new assets such as prompts and agents), software abstraction (i.e., native multi-agent support for cross-generation, sop-controllable software orchestration), executable intermediate representation (i.e., intent-rich graph representation for runtime optimization), and a fusion runtime that optimizes fmware for application level sla on heterogeneous computing resources such as gpus and npus.
报告人简介：
dayi lin is a principal researcher from centre for software excellence, huawei canada research institute, working on improving the productivity, quality and trustworthiness of conventional ai and foundation model powered software systems. he has published at several top-tier venues including tse, tosem, emse, icse and fse, and has attracted wide media coverage including kotaku, pc gamers, gamasutra, and national newspapers. he co-organized aiware 2024 (co-located with fse) and gas 2022-24 (co-located with icse).
https://lindayi.me/
张令明
报告题目：完全开源透明的代码大模型：构建和应用
报告摘要：
大语言模型（如gpt-4和claude-3.5等）在包括软件工程在内的各个领域下游应用中表现出色。近年来也涌现出了一批非常优秀的开源代码大模型。但绝大多数开源模型仅仅公布模型参数、而非训练数据集、过程和实现细节。长远来看这将阻碍代码大模型领域的进步和发展，比如模型的复现、增强、可解释性、去污、质量保证等问题。在本次报告中，我将讨论我们组近几年在构建完全开源、透明和自主的代码大模型方面的一些探索和经验（如starcoder2和magicoder等）。同时，我会简单探讨代码大模型对软件质量保证乃至软件工程带来的新机会和挑战，以及我们最近的一些尝试（比如ai自动软件工程师agentless）.
报告人简介：
张令明现任美国伊利诺伊大学香槟分校（uiuc）计算机系副教授。博士毕业于美国德州大学奥斯汀分校（ut austin），并分别于北京大学和南京大学获得硕士及学士学位。主要从事软件工程、程序语言以及机器学习的交叉研究。首次提出了一系列基于大模型的软件系统测试和修复技术（titanfuzz，alpharepair，chatrepair等）。并发布了一系列开源代码大模型及其验证数据集，累计下载量已超百万，相应技术也已经被多个业界最新代码大模型所采用（包括meta llama3.1）。近年来担任多个国际会议的程序委员会共同主席或者副主席（ase 2025、oopsla 2024等），受邀为多个会议和研讨会做主题报告（issta 2024、湘南会议等），并创办了第一届代码大模型workshop（llm4code 2024）。详细信息请见：http://lingming.cs.illinois.edu/
娄一翎
报告题目：llm-based agents for software engineering: current status and future opportunities
报告摘要：
the recent advance in large language models (llms) has shaped a new paradigm of ai agents, i.e., llm-based agents, which substantially extend the versatility and expertise of llms with the capabilities of perceiving and utilizing external resources and tools. llm-based agents have been applied and demonstrated remarkable effectiveness in various software engineering (se) tasks. this talk will first introduce the current status of llm-based agents for se, including their common design, multi-agent synergy, and human-agent collaboration. additionally, the talk will further discuss open challenges and future directions in llm-based agents for se.
报告人简介：
娄一翎，复旦大学计算机科学技术学院青年副研究员，分别于2016年和2021年在北京大学获得学士和博士学位，并曾于美国普渡大学计算机系任博后研究员。主要研究方向包括软件测试与分析、智能化软件开发等。在icse、fse、ase、issta、tse、tosem等软件工程国际高水平会议和期刊上发表论文30余篇，并多次获得acm sigsoft distinguished paper award和ieee tcse distinguished paper award。担任icse、fse、ase、issta等国际会议程序委员会委员和tse、tosem等国际期刊审稿人，以及多个会议和workshop的程序委员会共同主席（如llm4code 2024 和aiware 2025）。
thomas zimmermann
报告题目：aiware vs. empirical software engineering: the next frontier or final chapter?
报告摘要：
in this talk, we’ll explore how ai tools like gpt-4 are shaking up empirical software engineering by automating research replication. i’ll show experiments where gpt tackles research papers, nailing some tasks and stumbling on others. is this the next big thing or the final chapter in how we study software? let's find out!
报告人简介：
thomas zimmermann works on cutting-edge research and innovation in data science, machine learning, software engineering, and digital games. he has over 15 years of experience in the field, with more than 100 publications that have been cited over 25,000 times. his research mission is to empower software developers and organizations to build better software and services with ai. he is best known for his pioneering work on systematic mining of software repositories and his empirical studies of software development in industry. he has contributed to several microsoft products and tools, such as visual studio, github, and xbox. he is an acm fellow, an ieee fellow, recipient of the ieee tcse edward j. mccluskey technical achievement award, and co-editor in chief of the empirical software engineering journal.
jie m. zhang
panel嘉宾简介：
dr. jie m. zhang is a lecturer (assistant professor) of computer science at king’s college london, uk. before joining king’s she was a research fellow at university college london and a research consultant for meta. she got her phd degree at peking university in 2018. her main research interests are software testing, software engineering and ai/llms, and ai trustworthiness. she has published many papers in top-tier venues including iclr, icse, fse, ase, issta, tse, and tosem. she is a steering committee member of ieee icst and acm aiware. she is/was the general chair of aiware 2025, program co-chair of aiware 2024, internetware 2024, issta 2025 doctoral symposium, ase 2023 nier track, saner 2023 journal-first track, prdc 2023 fast abstract track, sbst 2021, mutation 2021&2020, and ase 2019 student research competition. over the last three years, she has been invited to give over 20 talks at conferences, universities, and it companies, including four keynote talks. she has been selected as the top-fifteen 2023 global chinese female young scholars in interdisciplinary ai. her research has won the 2022 transactions on software engineering best paper award and the iclr 2022 spotlight paper award.
04
论坛组织委员会简介
论坛主席：娄一翎
个人简介：
娄一翎，复旦大学计算机科学技术学院青年副研究员，分别于2016年和2021年在北京大学获得学士和博士学位，并曾于美国普渡大学计算机系任博后研究员。主要研究方向包括软件测试与分析、智能化软件开发等。在icse、fse、ase、issta、tse、tosem等软件工程国际高水平会议和期刊上发表论文30余篇，并多次获得acm sigsoft distinguished paper award和ieee tcse distinguished paper award。担任icse、fse、ase、issta等国际会议程序委员会委员和tse、tosem等国际期刊审稿人，以及多个会议和workshop的程序委员会共同主席（如llm4code 2024 和aiware 2025）。
论坛主席：thomas zimmermann
个人简介：
thomas zimmermann works on cutting-edge research and innovation in data science, machine learning, software engineering, and digital games. he has over 15 years of experience in the field, with more than 100 publications that have been cited over 25,000 times. his research mission is to empower software developers and organizations to build better software and services with ai. he is best known for his pioneering work on systematic mining of software repositories and his empirical studies of software development in industry. he has contributed to several microsoft products and tools, such as visual studio, github, and xbox. he is an acm fellow, an ieee fellow, recipient of the ieee tcse edward j. mccluskey technical achievement award, and co-editor in chief of the empirical software engineering journal.
论坛主席：谢涛
个人简介：
tao xie is a ccf fellow, director of ccf technical committee of system software (tcss), peking university chair professor, chair of the department of software science and engineering in the school of computer science at peking university, chief scientist of beijing institute of open source chip, deputy director of the key lab of high confidence software technologies, ministry of education. he was a full professor at the department of computer science, the university of illinois at urbana-champaign (uiuc), usa. he is a foreign member of academia europaea, and a fellow of acm, ieee, and aaas. he serves as the risc-v international ai/ml sig chair, risc-v ai computility ecosystem (race) committee chair, and risc-v summit china 2024 conference co-chair.
论坛主席：ahmed e. hassan
个人简介：
ahmed e. hassan is an acm fellow, ieee fellow, and an nserc steacie fellow. he is a laureate of the mustafa prize, commonly likened to a nobel, for founding the field of ai-augmented se/intelligent se and its worldwide impact on society and industry. he is the only individual globally to be recognized as a distinguished/influential educator by both of the world’s leading professional se societies, ieee-tcse and acm-sigsoft. he is a canada research chair (crc) in software analytics, and previously a nserc/blackberry software engineering chair. today he leads worldwide efforts to rethink software and se in the llm/fm era. see https://www.aiwarebootcamp.io,https://fmse.io, https://www.aiwareconf.org, and https://arxiv.org/abs/2402.15943 for more details. early systems and technologies developed by hassan and his teams are already integrated into products used by millions of users worldwide. his industrial experience includes architecting and scaling the global blackberry wireless platform, as well as working for ibm research at the prestigious almaden lab in usa, and the computer research lab of nortel networks. dr. hassan is the named inventor of patents in several jurisdictions around the world including the united states, europe, india, canada, and japan.
