← GeoAcademy System Administration College

AI Operations & Automation Department

The AI Operations & Automation Department teaches students to run production-grade AI/ML systems end to end through hands-on MLOps practice. The curriculum covers (1) building and versioning model registries with MLflow, (2) designing workflow orchestration with Airflow and Prefect, (3) implementing model-serving pipelines with TorchServe and TensorFlow Serving, and (4) instrumenting monitoring and running A/B tests for continuous improvement. Graduates are ready to step directly into deploying and operating reproducible, production-grade AI systems on their own.

⚙️ GeoAcademy System Administration College 🖥️ AI & Computing Sciences College 🔬 Basic Sciences College ⚡ Intelligent Geophysical Exploration College 🛢️ Resource & Energy Engineering College 🌍 Applied Geoscience Solutions College 💼 Economics, Policy & Strategy for the Future College 🎓 Education & Training Development College 🚀 Innovation & International Collaboration College

Infrastructure & Cloud Operations Department Database & Knowledge Systems Department AI Operations & Automation Department Cybersecurity & Compliance Department Application & Simulation Operations Center Meta-Governance & Interface Center Software Engineering & Development Center

🔑

⭐ Shannon

Chair

💡 Information Theory & AI Operations

Researchers (real GeoVerse Lab members) 17

🔑

1861–1919

Henry Gantt

Researcher

💡 Invented the Gantt chart (1910s), the graphical dependent-task scheduling method that is the direct conceptual ancestor of the DAG-based workflow schedulers (Airflow, Prefect, Luigi) used to orchestrate ML pipelines

🔑

1876–1937

William Sealy Gosset

Researcher

💡 Developed Student's t-distribution and t-test (1908, under pen name "Student") for drawing statistically valid conclusions from small numbers of trials — the same regime MLOps experiment-tracking tools face when comparing few costly training runs

🔑

1931–2022

Frederick Brooks

Researcher

💡 Managed IBM System/360 and OS/360 development and wrote "The Mythical Man-Month" (1975), the founding text on large-scale software project, version, and release management that underlies modern model-registry and release-governance practice

🔑

1924–2017

Charles Bachman

Researcher

💡 Built the Integrated Data Store (IDS, 1963), one of the first database management systems, the network data model, and Bachman diagrams for visualizing data structure and lineage — foundational to modern dataset versioning and data-provenance tracking (DVC-style tools)

🔑

1887–1979

Waloddi Weibull

Researcher

💡 Developed the Weibull distribution (1951), the standard statistical model of component/system time-to-failure, now used to model server, GPU, and inference-endpoint reliability and to set SLA/uptime targets for model-serving infrastructure

🔑

1901–1972

Ludwig von Bertalanffy

Researcher

💡 Founded General Systems Theory (1940s-1968), the interdisciplinary framework for understanding complex systems of interacting components as open systems exchanging information/energy with their environment — the theoretical basis for holistic monitoring/observability of production ML systems

🔑

1891–1967

Walter Shewhart

Researcher

💡 Invented statistical process control and the control chart (1924) for detecting when a manufacturing process has shifted out of its stable statistical distribution — the direct statistical ancestor of automated data-drift and concept-drift detection in production ML monitoring

🔑

1878–1929

Agner Krarup Erlang

Researcher

💡 Founded queueing theory and teletraffic engineering (Erlang B/C formulas, 1909-1917) for sizing telephone exchange capacity against fluctuating call volume — the mathematical foundation for autoscaling policies that size model-serving replica pools against fluctuating inference traffic

🧑‍🔬

🔑

1932–1982

Gwilym Jenkins

Researcher

💡 Co-developed (with George Box) the Box-Jenkins ARIMA methodology (1970) for time-series forecasting and model identification — the standard method for forecasting resource usage, request volume, and system-load trends used in AIOps capacity planning

🔑

1931–2021

Jonas Mockus

Researcher

💡 Founded Bayesian global optimization theory (from the 1970s), the surrogate-model-based method for efficiently optimizing expensive-to-evaluate black-box functions — the core algorithmic basis of modern hyperparameter-optimization and AutoML systems

🔑

1884–1965

Corrado Gini

Researcher

💡 Developed the Gini coefficient/index of statistical dispersion (1912), which became the standard impurity/importance criterion (Gini impurity, mean decrease in Gini) used to explain and rank feature contributions in decision-tree and ensemble models deployed in production ML monitoring dashboards

🔑

1885–1970

Nicolas Minorsky

Researcher

💡 First rigorously analyzed and proposed the PID (proportional-integral-derivative) feedback controller (1922) for automatic steering of US Navy ships — the foundational feedback-control algorithm behind automated self-healing and closed-loop remediation systems in AIOps

🧑‍🔬

🔑

1895–1980

Egon Pearson

Researcher

💡 Co-developed (with Jerzy Neyman) the Neyman-Pearson framework of statistical hypothesis testing (1928-1933), including the concepts of significance level, statistical power, and Type I/II error — the formal statistical basis for online A/B testing and controlled-experiment model-performance comparison

🔑

1907–1983

Maurice Kendall

Researcher

💡 Developed Kendall's tau rank correlation coefficient (1938) and related rank-statistics methods for measuring association between variables — a standard tool for feature-correlation analysis, redundancy detection, and feature selection used when building and curating feature stores

🧑‍🔬

🔑

1916–1975

Christopher Strachey

Researcher

💡 Pioneered computer time-sharing (proposing multi-user shared computer access in 1959) and co-founded denotational semantics — foundational concepts of process/resource isolation on shared hardware that underlie today's containerization and virtualization technologies (Docker, Kubernetes) used to isolate model-serving workloads

🧑‍🔬

🔑

1927–2010

Watts Humphrey

Researcher

💡 Founded the Capability Maturity Model (CMM) and Personal/Team Software Process (PSP/TSP) at the SEI (1986-2000s), the disciplined, measured, and continuously-improved software-process framework that underlies modern continuous-integration/continuous-deployment (CI/CD) practice for ML systems

🧑‍🔬

🔑

1937–1995

Barry Turner

Researcher

💡 Founded "disaster incubation theory" in "Man-Made Disasters" (1978), showing that large organizational failures accumulate silently through unnoticed small warning signs before erupting — the sociological foundation for blameless postmortem and incident-review culture in SRE/AIOps practice