Postdoctoral AI Scientist

Palo Alto, California | 12-24Month Appointment

About AIVista

NTT DATA AIVista, Inc., a wholly owned subsidiary of NTT DATA, develops AI products for enterprises operating in complex regulatory environments. Based in Palo Alto, we combine deep AI product expertise with NTT DATA's industry knowledge and systems-integration experience, working with NTT companies to deploy solutions for enterprise clients.

The Opportunity

AIVista is seeking a Postdoctoral AI Scientist to conduct research on foundational problems at the intersection of machine learning, knowledge representation, formal methods, and AI agents for enterprise systems.

You will work alongside scientists and engineers to define and pursue ambitious research questions, develop new methods and systems, and evaluate them against complex, production-scale problems and data. You will conduct research that advances the state of the art while creating a path toward performant and trustworthy AI systems that can operate reliably in real enterprise environments.

This is an an initial 12-month appointment with the option to extend it to 24 months postdoctoral research appointment designed for researchers who have completed their PhD and want to continue developing a strong research agenda while working closely with real-world AI systems. Pursuing publication of research at top-tier conferences and journals is highly encouraged, and projects are selected to support both rigorous scientific contribution and practical relevance.

Science Focus

Neurosymbolic methods and trustworthy reasoning: Combine learning-based and symbolic techniques to improve reliability, transparency, and alignment with domain constraints.

Knowledge representation and semantic AI: Develop methods that help AI systems organize, connect, and reason over complex enterprise information.

Adaptive and model-agnostic AI systems: Explore flexible approaches that integrate models, tools, context, and memory across diverse tasks and environments.

Scalable domain model development

Develop scalable methods for building domain-specialized models, spanning automated generation of training and agentic environments, mid-training on domain data, post-training through fine-tuning and reinforcement learning, and rigorous domain-grounded evaluation.

Learning from feedback: Develop methods that enable AI systems to improve safely and effectively from human and operational feedback.

Evaluation, verification, and interpretability: Advance rigorous approaches to assessing AI reliability, robustness, safety, and transparency.

AI for complex workflows: Explore how intelligent systems can support and improve multi-step enterprise processes while maintaining appropriate oversight and control.

What You'll Do

Define and independently drive research questions across one or more of AIVista's science focus areas.

Develop novel algorithms, models, methods, prototypes, and system designs and evaluate their effectiveness through rigorous experimentation.

Design experiments and benchmarks that measure correctness, robustness, calibration, privacy, latency, cost, and process outcomes.

Work with scientists, engineers, and domain experts to translate enterprise data, policies, feedback, and operational constraints into research problems, research artifacts, and deployable systems.

Develop inspectable AI systems and analyses that connect model and agent decisions to evidence, rules, and outcomes.

Move promising research from early experimentation through simulation, controlled evaluation, and potential application in production-scale AI systems.

Communicate research findings internally and externally and contribute high-quality research to leading conferences, journals, and the broader scientific community.

Collaborate with and contribute to AIVista's broader research agenda while developing an independent body of research during the postdoctoral appointment.

Qualifications

PhD in Computer Science, Machine Learning, Artificial Intelligence, Statistics, Applied Mathematics, Electrical or Computer Engineering, Operations Research, or a closely related technical field. The PhD must be completed prior to the start of the appointment.