Builders & open source
Ruchir Puri
Chief Scientist, IBM Research; IBM Fellow; Vice President, IBM Technology & Technical Community, IBM
- Chief Scientist, IBM Research; IBM Fellow; Vice President, IBM Technology & Technical Community, IBM Research1
Ruchir Puri is the Chief Scientist of IBM Research, an IBM Fellow, and Vice President of IBM Technology and Technical Community, according to the IBM Research profile. The profile notes that Puri led IBM Watson as its CTO and Chief Architect from 2016 to 2019. Puri is the first author of the 2021 paper introducing CodeNet, an AI-for-code dataset whose IBM/Project_CodeNet repository (Apache-2.0) provides processing tools and a download link for the full dataset, and a co-author of IBM's Granite Code models paper, whose models were released under the Apache 2.0 license. Puri also appears among the authors of ITBench, a benchmark for AI agents on IT automation tasks.12345
At a glance
- Current role
- Chief Scientist, IBM Research; IBM Fellow; Vice President, IBM Technology & Technical Community
- Company
- IBM
- Sector
- Company
- Education
- Columbia, Stanford
- Work listed
- 3
- Honors
- 2
- Sources cited
- 8
- Profile reviewed
Career
Puri has held technical, research and engineering leadership roles across IBM's AI and research businesses. From 2016 to 2019 Puri led IBM Watson as its CTO and Chief Architect, and Puri is now Chief Scientist of IBM Research, an IBM Fellow and Vice President of IBM Technology and Technical Community.1
Outside IBM, Puri has been an adjunct professor at Columbia University and a visiting scientist at Stanford University, held the John von Neumann Chair at the Institute of Discrete Mathematics at the University of Bonn, and has served as an ACM Distinguished Speaker and an IEEE Distinguished Lecturer.1
Research
Puri's recent research applies language models to software and hardware engineering. Puri was first author of the 2021 CodeNet dataset paper and co-authored IBM's Granite Code models (2024), the Qiskit Code Assistant for generating quantum computing code (2024) and a study of customizing a language model for VHDL design of high-performance microprocessors (2025).2467
Work on AI agents includes ITBench (ICML 2025), a benchmark of IT automation tasks, and a 2026 paper on how far general-purpose coding agents can go in optimizing hardware through high-level synthesis.518
Work and contributions
IBM Granite Code models Open weights
Listed among the authors of "Granite Code Models: A Family of Open Foundation Models for Code Intelligence" (2024), which introduces decoder-only code models from 3 to 34 billion parameters and states that all Granite Code models are released under an Apache 2.0 license.4
Project CodeNet Open data
First author of "CodeNet: A Large-Scale AI for Code Dataset for Learning a Diversity of Coding Tasks" (2021). The IBM/Project_CodeNet repository, whose code is under the Apache-2.0 license, provides tools for the dataset and download instructions for the full dataset.23
ITBench Published research
Listed among the authors of "ITBench: Evaluating AI Agents across Diverse Real-World IT Automation Tasks" (2025).5
Timeline
- Became CTO and Chief Architect of IBM Watson, serving until 2019.1
- Released the CodeNet AI-for-code dataset paper as first author.2
- Co-authored "Granite Code Models - A Family of Open Foundation Models for Code Intelligence".4
- Co-authored the Qiskit Code Assistant paper on training LLMs to generate quantum computing code.6
- Co-authored ITBench, a benchmark for AI agents on IT automation tasks.5
- Co-authored a study on customizing a language model for VHDL microprocessor design.7
- Co-authored a paper on general-purpose coding agents for hardware optimization via high-level synthesis.8
Other documented roles
Roles the sources document that are not confirmed as current.
- CTO and Chief Architect (2016–2019), IBM Watson1
Honors and recognition
Where they work
- IBM 29 people in the wiki
Related people
- Marina DanilevskyAlso at IBM
- Sriram RaghavanAlso at IBM
- Ranit AharonovAlso at IBM
- Kate SouleAlso at IBM
- Avi SilAlso at IBM
- Leo GaoLanguage models · Evaluation and benchmarks
- Jesse DodgeEvaluation and benchmarks · Datasets and data
- Swabha SwayamdiptaEvaluation and benchmarks · Datasets and data
- Eugene YanLanguage models · Evaluation and benchmarks
- Rohan TaoriLanguage models · Evaluation and benchmarks
Suggest a correction or removal
Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.
References
8 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026
- Ruchir Puri — IBM Research
- CodeNet: A Large-Scale AI for Code Dataset for Learning a Diversity of Coding Tasks (arXiv 2105.12655)
- IBM/Project_CodeNet README (dataset download) and license (Apache-2.0)
- Granite Code Models: A Family of Open Foundation Models for Code Intelligence (arXiv 2405.04324)
- ITBench: Evaluating AI Agents across Diverse Real-World IT Automation Tasks (arXiv 2502.05352)
- Qiskit Code Assistant: Training LLMs for generating Quantum Computing Code (arXiv 2405.19495)
- Customizing a Large Language Model for VHDL Design of High-Performance Microprocessors (arXiv 2505.09610)
- Agent Factories for High Level Synthesis: How Far Can General-Purpose Coding Agents Go in Hardware Optimization? (arXiv 2603.25719)