Sean (Xiang) Ren is the co-founder and CEO of Sahara AI, a decentralized AI blockchain platform. He serves as an Associate Professor and the Andrew and Erna Viterbi Early Career Chair in Computer Science at the University of Southern California (USC), where he directs the Intelligence and Knowledge Discovery (INK) Research Lab. ^[1]

Education

Sean Ren earned his Bachelor of Engineering degree in Computer Science from Zhejiang University. He later pursued his doctoral studies at the University of Illinois Urbana-Champaign (UIUC), where he completed his Ph.D. in Computer Science. During his PhD work, he also spent time with the Natural Language Processing (NLP) group and the SNAP group at Stanford University. ^{[2] [16]}

Career

Ren began his academic career at the University of Southern California (USC), joining the Computer Science Department in December 2017. He was appointed as an Associate Professor and the Andrew and Erna Viterbi Early Career Chair at USC in September 2022. He also holds an appointment as a Research Team Leader at USC's Information Sciences Institute (ISI) and directs the Intelligence and Knowledge Discovery (INK) Research Lab, which focuses on advancing NLP and machine learning.

Beyond his academic roles, Ren has held positions in industry, including serving as a Data Science Advisor at Snap Inc. and as a Visiting Research Scientist at the Allen Institute for AI (AI2) from April 2022 to March 2024. In May 2023, he co-founded Sahara AI, a decentralized AI blockchain platform, where he serves as CEO. The company aims to foster the development of novel AI for fair, transparent, and universal access to global knowledge capital, focusing on areas like provenance for AI intellectual property and user-owned, portable agents. Sahara AI has engaged in partnerships for on-chain co-ownership and revenue sharing for data contributors, and its token, $SAHARA, has been listed on exchanges.

Ren is also actively involved in the broader academic community, serving as the Information Director for ACM SIGKDD and the annual KDD Conference since September 2018. He has organized numerous workshops at major conferences, including the Workshop on Commonsense Representation and Reasoning (CSRR), Federated Learning for NLP (FL4NLP), and Deep Learning for Low-Resource NLP (DeepLo). ^{[1] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [2]}

Research

Ren's research primarily focuses on building generalizable natural language processing (NLP) systems capable of handling diverse language tasks and situations, with an emphasis on broadening the scope of model generality. His work involves developing new algorithms and datasets to reduce the cost of building and maintaining NLP models, equipping AI models with common sense, and enhancing model transparency and reliability to build user trust.

The INK Research Lab, led by Ren, concentrates on several key areas:

• Creating evaluation methods and datasets to assess state-of-the-art NLP systems in various human reasoning scenarios.
• Building novel learning algorithms and model architectures to augment NLP systems with commonsense and factual knowledge.
• Developing graph neural network methods for relational reasoning.
• Verifying and enhancing the robustness of NLP models.

His research also explores extracting machine-actionable knowledge from natural language data, performing neural-symbolic knowledge reasoning for intelligent applications, and learning from human explanations and instructions. His PhD work on label-efficient NLP is summarized in the book "Mining Structures of Factual Knowledge from Text: An Effort-Light Approach".

Ren's research has received funding from various organizations, including the National Science Foundation (NSF CAREER award), DARPA (MCS, KMASS, INCAS, SCORE, GAILA, SAIL-ON), IARPA (HIATUS, BETTER), and industry partners such as Google, Amazon, Meta, JP Morgan, Adobe, Sony, and Snapchat. ^{[2] [7]}

Awards and Recognition

Sean Ren has received numerous awards and recognitions throughout his career:

• 2024: ACM WSDM Test of Time Award.
• 2023: Samsung AI Researcher of the Year.
• 2023: MIT Technology Review Innovators Under 35 (Asia Pacific).
• 2023: Okawa Research Grant.
• 2023: ACL Outstanding Paper Award for "SCOTT: Self-Consistent Chain-of-Thought Distillation".
• 2022: NAACL Outstanding Paper Award for "NewsEdits".
• 2022: Google Research Scholar.
• 2021: NSF CAREER Award.
• 2021: Amazon Secure and Trusted ML Grant.
• 2021: Best Paper Award at TrustNLP @NAACL.
• 2020: Sony Faculty Innovation Award.
• 2020: Best Paper Award Runner-up at The Web Conference for "NERO: A Neural Rule Grounding Framework for Label Efficient Relation Extraction".
• 2019: Forbes’ Asia 30 Under 30 in Healthcare & Science.
• 2019: JP Morgan AI Research Award.
• 2019: Amazon Faculty Research Award.
• 2018: Google Faculty Research Award.
• 2018: ACM SIGKDD Doctoral Dissertation Award.
• 2017: David J. Kuck Outstanding Thesis Award.
• 2016: Google PhD Fellowship. ^{[2] [8] [9] [10] [11] [12] [13] [14]}

Teaching

Ren has taught several courses at USC, focusing on deep learning, natural language processing, and knowledge extraction. These include:

• Deep Learning and Its Applications (CSCI 566)
• Representation Learning for NLP
• Advanced Topics in Representation Learning for NLP (CSCI 699)
• Machine Learning for Knowledge Extraction and Reasoning (CSCI 699)
• Introduction to Information Extraction.

Ren is actively involved in academic service, serving as an Area Chair or Senior Program Committee member for major AI and NLP conferences such as ACL, EMNLP, NeurIPS, ICML, AAAI, and ICLR. He has also contributed to K-12 outreach initiatives, giving guest lectures and participating in project panels to explain AI applications, including combating online hate speech, to high school students. ^{[2] [15]}

Interviews

Sean Ren on AI and Blockchain #01

In an episode of The Tech Optimist podcast, Sean Ren, co-founder and CEO of Sahara AI, outlined the project’s framework and its intended role at the intersection of artificial intelligence and blockchain technologies. According to Ren, Sahara AI aims to develop a decentralized infrastructure that enables users to manage, retain, and trade AI-related data under conditions designed to support transparency and privacy.

Ren described the platform’s technical structure as consisting of three main layers. The base is a custom Layer 1 blockchain, created to record ownership and provenance of datasets, models, and applications associated with AI workflows. Built on top of this is a decentralized execution layer, which facilitates the training and deployment of AI models using mechanisms such as Low-Rank Adapters (LoRA), intended to maintain data privacy. The third layer includes tools and marketplaces for tasks such as data labeling, agent configuration, and model access management.

During the interview, Ren stated that Sahara AI had recently released its testnet, which is integrated with its application marketplace. He also mentioned plans for the launch of a mainnet and the introduction of additional tools by the end of the year. The company disclosed a funding round totaling $43 million, with participation from investment entities such as Binance Labs, Pantera Capital, and Polychain Capital. According to Ren, these resources will be allocated toward workforce expansion, ecosystem development, and outreach initiatives.

Among the topics discussed were the relevance of data and model traceability in AI systems, and the use of blockchain to register and manage such provenance. Ren emphasized that implementing these mechanisms could support more structured oversight of AI components in decentralized environments. He also pointed to decentralized LoRA as a method for users to fine-tune models locally while minimizing exposure of sensitive data.

Ren addressed the lack of integration between blockchain and AI communities, noting that Sahara AI is engaged in efforts to promote knowledge exchange and cooperative development across both fields. He also provided context on his academic and professional background, which includes research in natural language processing and prior experience in building AI systems. According to Ren, this combination of expertise contributes to the project's structural approach to technical implementation and data governance. ^[18]

Sean Ren on Decentralized AI Infrastructure #02

On February 18, 2025, Sean Ren, CEO and co-founder of Sahara AI, appeared on the Cryptonews Spotlight podcast to discuss the development of decentralized artificial intelligence systems and the application of blockchain technology in AI workflows. The conversation addressed the emergence of open-source AI models and their potential implications for existing centralized infrastructures.

Ren outlined how publicly available models such as DeepSeek offer an alternative to proprietary systems, allowing independent developers to modify and deploy models under specific resource conditions. He described Sahara AI’s role in facilitating transparent attribution of data, compute, and model contributions through blockchain-based tracking mechanisms.

The interview also introduced the concept of knowledge agents (KAs), AI systems that operate using user-specific or organization-specific data to automate tasks. According to Ren, these agents are designed to function within defined contexts by integrating personal data, potentially improving operational efficiency in various domains.

In addition, Ren explained the structure of Sahara AI’s marketplace, which supports the distribution and monetization of datasets, models, and computing services. The platform applies programmable contracts to allocate compensation based on usage and performance, aiming to align incentives across contributors.

The discussion included references to institutional collaborations involving major technology companies and highlighted initiatives designed to increase engagement from developers and end-users. These efforts include the release of new tools and ecosystem-wide participation programs.

Ren presented Sahara AI’s approach as one that applies decentralized infrastructure to address challenges in data ownership, intellectual property management, and contributor compensation in AI development. ^[17]