3 Former MIT Researchers at IBM Push AI and Quantum Deployment

3 articles · Updated · MIT News · Sep 2

Three former MIT researchers now at IBM — Zhang-Wei Hong, Irene Ko and Srinivasan Arunachalam — are turning academic work into deployable AI and quantum systems through the MIT-IBM Computing Research Lab.
Hong, who completed his PhD in 2025, is building reinforcement-learning infrastructure for IBM’s enterprise agent framework, including test-time training and online self-improving models for tasks such as chart reading and database tool calling.
Ko, who joined IBM Research after graduating in 2024, is developing trustworthy-AI tools including a lightweight vLLM Hook plugin that reads internal model signals to flag prompt injection and hallucination while cutting inference costs.
Arunachalam, an MIT postdoc from 2018 who later moved to IBM, has focused on near-term quantum machine learning, with work on Hamiltonian learning and quantum kernels aimed at provable advantages under real hardware constraints.
The lab’s role across all three careers was to connect theory-heavy MIT research with IBM product needs, helping move ideas from novelty and rigor toward business-relevant applications.