My PhD at Oakland University's Complex Systems Laboratory studies how leaders should divide attention between an AI recommender and their own experts. An engineering portfolio faces far more stage-gate decisions than leadership can carefully review, so my dissertation builds a risk-aware decision-support framework: routine gate decisions proceed on the AI recommendation, while the small set of uncertain, high-consequence decisions is routed to human reviewers — under an explicit limit on worst-case portfolio losses. Coursework is complete; the topic proposal is scheduled for fall 2026, with expected completion in April 2027. The work has also produced a U.S. patent application on multi-modal explanation generation, a published conference paper, and a systematic review under journal review.
Dissertation · Proposal — Fall 2026
Risk-Aware Human–AI Decision Support for Selective Review in Engineering Stage-Gate Portfolios
Engineering leaders cannot personally review every AI-supported gate decision in a large portfolio — and they should not have to. This dissertation develops a decision-support policy that allocates a limited budget of expert reviews across many stage-gate decisions: routine calls proceed on the AI recommendation, while the decisions where review adds the most value — the ambiguous, the novel, the high-stakes — are escalated to humans, all subject to a cap on worst-case portfolio losses.
- Human–AI teaming
- Selective review
- Risk-aware RL
- Stage-gate governance
- XAI
U.S. Patent Application · Filed May 2026
System and Method for Multi-Modal Explanation Generation with Cross-Modal Consistency Verification for AI-Driven Engineering Project Decision Support
A method for generating decision-grade explanations across textual, visual, and quantitative modalities — with a cross-modal consistency check that flags contradictions between channels before an engineering manager acts on a recommendation. Filed as a U.S. provisional application with the USPTO.
- USPTO 64/067,534
- Filed 16 May 2026
- Sole inventor
Paper · Published — IEOM 2026
Multi-modal Explanations for Engineering Project Decision Support
Introduces the Tri-Modal Explanation Architecture (TMEA): three coordinated outputs — a Gantt schedule overlay, a KPI partial-derivative panel, and a structured decision rationale — from a single decision-tree surrogate of a PPO policy, with a cross-modal consistency mechanism that withholds misaligned explanations. Published at the 11th North American IEOM Conference, Milwaukee, June 2026. Open access via Zenodo (DOI).
- XAI
- Multi-modal
- Decision support
Paper · Preprint — SSRN 2026 · Under review
Deep Reinforcement Learning for Engineering Project Management: A Systematic Review
A systematic review of how deep reinforcement learning has been applied to engineering project and portfolio decisions — the methods used, the strength of the evidence, and the open gaps where RL meets real project governance. Posted on SSRN and under review at Engineering Applications of Artificial Intelligence. Read on SSRN.
- Systematic review
- Deep RL
- Project management
Master's Thesis · TU Berlin 2015 · Published — Zenodo
Geometric design of 3-dimensional backward curved impeller blade with constant blade thickness
M.Sc. thesis in Global Production Engineering at Technische Universität Berlin, supervised by Prof. Dr.-Ing. Paul Uwe Thamsen and carried out with Konrad Reitz Ventilatoren. A CFD-driven design study (ANSYS CFX) of a backward-curved centrifugal-fan impeller blade, optimizing blade inlet angle, leading-edge radius, twist, and diameter factor to improve fan efficiency. Open access via Zenodo (DOI).
- Centrifugal fans
- CFD
- Impeller design
Paper · In preparation
A Latent Trust Estimation Model for Adaptive Explanation Delivery in Human–AI Decision Support Systems
Introduces a latent-variable model that infers a user's evolving trust in an AI recommender from interaction signals and adapts the depth, modality, and timing of explanations in real time — closing the loop between explainability and human–AI collaboration in engineering project management.
- Trust modeling
- Adaptive XAI
- Human–AI teaming
Scholarly Service · Peer Reviewer — Q1 Journals
Peer Reviewer — AI, Decision Support & Engineering Management
Reviewer for leading Elsevier and IEEE journals, including Engineering Applications of Artificial Intelligence and MethodsX, with completed reviews of reinforcement-learning and optimization manuscripts. Registered reviewer for Expert Systems with Applications, Computers & Industrial Engineering, Knowledge-Based Systems, Decision Support Systems, Applied Soft Computing, International Journal of Project Management, IEEE Transactions on Engineering Management, and other Q1/Q2 journals in AI and engineering management.
Invited Speaker · ICIST 2026 — Tashkent
Trustworthy AI for High-Stakes Engineering Decisions: Risk-Aware Reinforcement Learning with Selective Human Escalation
Invited talk at the International Conference on Intelligent Infrastructure and Smart Technologies (ICIST 2026), hosted by Tashkent University of Information Technologies, Uzbekistan — delivered online on 16 October 2026.
- Invited talk
- Trustworthy AI
- Human oversight
Keynote Speaker · ISASF 2026 — Sydney
Intelligent Decisions for a Sustainable Future: Risk-Aware AI and Human Oversight in Autonomous Engineering Systems
Keynote at the IEEE International Conference on Intelligent Sensing and Autonomous Systems for Sustainable Future (ISASF 2026), Western Sydney University, Sydney, Australia — 19 November 2026 (hybrid).
- Keynote
- IEEE
- Sustainable AI systems
Lab
Complex Systems Laboratory
Active member of Oakland University's Complex Systems Lab, collaborating on modeling and simulation of engineered systems.
- Systems modeling
- Simulation