Kazuhira (Kazu) Okumoto, Ph.D.
Expert Profile

Kazuhira (Kazu) Okumoto, Ph.D.

Founder & CEO, Sakura Software Solutions

Software reliability engineering pioneer and creator of predictive quality platforms.

Kazuhira (Kazu) Okumoto, Ph.D. is Founder and CEO of Sakura Software Solutions and a longtime contributor to software reliability engineering, predictive software quality, and system reliability modeling.

His work connects foundational reliability research with operational tools including STAR for software defect prediction, EVR analysis, residual defect forecasting, and FUSION for integrated software and hardware reliability assessment.

Authority signals

Former Bell Labs scientist with 40 years of software quality and reliability work.

Distinguished Member of Technical Staff at Bell Labs, the organization’s highest research rank.

Co-creator of the widely cited Goel-Okumoto NHPP software reliability model.

Co-author of Software Reliability: Measurement, Prediction, Application.

Former Assistant Professor at Rutgers University.

Principal Investigator on NSF-funded research and innovation programs.

Creator of the STAR and FUSION predictive reliability platforms.

Career timeline

1975 - 1979

Ph.D., Syracuse University

  • Funded by Rome Air Development Center (RADC).
  • Co-developed the Goel-Okumoto NHPP software reliability model.
1979 - 1980

Assistant Professor, Rutgers University

  • Secured NSF funding for early software reliability engineering research.
1980 - 2020

Bell Labs / AT&T / Lucent / Alcatel-Lucent / Nokia Bell Labs

  • Served as Distinguished Member of Technical Staff.
  • Co-authored Software Reliability: Measurement, Prediction, Application.
  • Led advanced reliability research and U.S. Navy development programs.
  • Developed BRACE, the original precursor to the STAR predictive platform.
  • Directed international technical programs in Japan and Ireland.
2020 - Present

Founder & CEO, Sakura Software Solutions

  • Founded 3S to operationalize predictive reliability engineering.
  • Created the STAR and FUSION platforms.
  • Serves as Principal Investigator on NSF-funded innovation programs.

Expertise areas

Software defect predictionSoftware reliability growth modelsEVR and backlog stabilityResidual defect forecastingRelease readiness analyticsIntegrated software-hardware reliabilityReliability block diagram analysisPredictive QA decision support

Selected authored publications

These publication pages connect the author entity to the supporting papers, white papers, conference presentations, and book chapters hosted by 3S.

RIDT2026

RIDT: Reliability-State-Aware Performance Analytics for Open RAN Systems

RIDT: Reliability-State-Aware Performance Analytics for Open RAN Systems is a white paper from 3S White Paper Series focused on RIDT.

View citation page
ADM2026

ADM: An Agile Development Manager for Early Defect Prediction in Small Software Teams

A 2026 ISSAT RQD conference paper and presentation introducing ADM, an Agile Development Manager model for early defect prediction in small software teams, presented at the Atlanta conference.

View citation page
FUSION2026

FUSION: Predictive Reliability–Performance Analytics for Open RAN Systems

A white paper applying the FUSION reliability framework to Open RAN systems, delivering predictive reliability and performance analytics for modern disaggregated radio access network architectures.

View citation page
STAR / FUSION2026

From Residual Defects to Operational Failures: A Two-Stage Mapping Framework for Predictive Operational Risk

A white paper presenting a two-stage mapping framework that links residual software defects to operational failures, enabling predictive operational risk assessment across STAR and FUSION tool outputs.

View citation page
STAR / ADM2026

ADM: An Agile Development Manager for Early Defect Prediction in Small Software Teams

A white paper introducing ADM, an Agile Development Manager approach for early defect prediction and release-risk visibility in small software teams.

View citation page
EVR2026

Introducing A Modern Software Metric, EVR, for Predictability, Stability & High-Quality Delivery

A white paper introducing Escape Velocity Rate (EVR) as a software quality metric for backlog stability, predictability, and high-quality delivery.

View citation page
ROI2026

Quantifying the Business Value of Predictable Software Delivery: A Two-Year ROI Analysis of 3S Reliability Tools

A two-year ROI analysis explaining how predictive reliability tools can quantify business value, reduce quality risk, and support predictable software delivery.

View citation page
FUSION2026

An NSF-awarded Digital Engineering Tool for Integrated System Reliability Assessment

A RAMS 2026 presentation on FUSION, an NSF-awarded digital engineering tool for integrated software and hardware reliability assessment.

View citation page

Connected pages

STAR

Predictive software quality platform for defect prediction, EVR, and release readiness.

FUSION

Integrated software and hardware reliability forecasting platform.

Research & Publications

Canonical hub for publications, conference papers, and citation pages.

Brand & Citation Kit

Official citation targets, descriptions, and profile references.