NTT DATA AIVista
Postdoctoral AI Scientist
$150K
San Francisco Bay Area · Hybrid · Internship · Posted Oct 23, 2026
- Education
- PhD in Computer Science, Machine Learning, Artificial Intelligence, Statistics, Applied Mathematics, Electrical or Computer Engineering, Operations Research, or a closely related technical field; must be completed before the appointment starts.
About the job
Conduct independent research on foundational AI methods and develop, evaluate, and potentially transition trustworthy AI systems for complex enterprise environments.
What you'll do
- Define and lead research questions; develop algorithms, models, prototypes, and system designs, and evaluate them through rigorous experiments.
- Design benchmarks and assess AI systems for correctness, robustness, safety, privacy, latency, cost, and other outcomes.
- Collaborate with research, engineering, and domain teams to translate enterprise needs into research and deployable systems; communicate findings and contribute to publications.
What they ask for
- Completed PhD before appointment start in a listed technical field or a closely related technical field.
- Demonstrated research record, such as peer-reviewed publications, significant projects, or comparable scientific contributions.
- Strong foundations in machine learning, algorithms, statistical methods, experimental design, and empirical research; deep research experience in at least one listed AI area.
- Proficiency in Python and experience implementing and evaluating machine learning or AI systems.
Nice to have
- Additional research experience in areas such as formal verification, knowledge graphs, domain-model training, agent systems, privacy-preserving learning, or process mining.
- Research involving AI deployed in regulated, safety-sensitive, or other high-stakes environments.
Pay
$150,000 USD annual base salary
Tools
Python
AI focus
Research and develop trustworthy enterprise AI, including neurosymbolic reasoning, knowledge representation, domain-specialized models, AI agents, learning from feedback, and evaluation, verification, and interpretability.
Model-written summary · the original posting is authoritative