Richard Martin Flores

Defense Analyst / Data Scientist

r.m.flores@outlook.com

Department of War

NASA Systems Analysis and Concepts Directorate

NASA Space Mission Analysis Branch

NASA Langley Research Center

Blog

Publications

Linkedin

“The people who are crazy enough to think they can change the world are the ones who do.” ~ Steve Jobs

I have always been driven by a single question: How do we build better models of the world? My career has been a journey through the different ways we answer that starting with the deterministic foundations of Physics and evolving into the high-stakes world of Operations Research and Data Science.

During my three years as a Data Scientist at NASA, I focused on the intersection of large-scale datasets and mission-critical accuracy. Working in an environment where the margin for error is near zero taught me how to scale analytical models that are both robust and actionable. It wasn't just about the data; it was about the systems-level thinking required to support complex, multi-year objectives.

Currently I serve as a Modeling & Simulation Specialist. My work is focused on the architecture of decision-support systems. By leveraging my background in physics and data analytics, I develop high-fidelity simulations that help stakeholders navigate uncertainty and forecast outcomes in increasingly volatile environments. I view simulation not just as a tool, but as a laboratory for strategy.

Beyond my professional role, I am currently a PhD student in Data Science. My research is dedicated to the evolution of Artificial General Intelligence (AGI) and its potential to redefine how we approach complex systems. I believe the next frontier of modeling lies in the bridge between human-centric game theory and autonomous agent-based systems.

Core strengths

• Research domains: Neuromorphic computing; cooperative game theory; multi-agent systems; continual and meta learning; credit assignment and alignment.
• ML & AI methods: Deep learning, Bayesian inference, reinforcement learning (policy gradients, off-policy), causal discovery, probabilistic programming.
• Systems & hardware: Neuromorphic accelerators, model-hardware co-design, GPU/TPU pipelines, distributed model serving, on-edge continual learning.
• Engineering stack: Python, PyTorch, JAX, TensorFlow; AWS, Snowflake; dbt; SQL; containerization and orchestration (Docker, Kubernetes).
• Domain expertise: Aerospace systems analysis, financial modeling and risk quantification, explainability, safety-critical verification.

Research

1. Cooperative game theoretical credit assignment for multi-agent learning
• Formalize Shapley-inspired credit allocation methods that scale to temporally extended tasks.
• Apply to multi-vehicle coordination and distributed resource allocation, optimizing both team reward and individual incentive compatibility.
2. Modular, composable AGI curricula with provable transfer bounds
• Design architectures of specialized modules (perception, planning, memory) and protocols for gating and composition.
• Prove and empirically validate transfer generalization bounds under curriculum distributions, measure catastrophic forgetting.
3. Causal representation learning for robust generalization
• Learn disentangled causal factors using interventional-style data augmentation and invariance objectives.
• Integrate learned causal units into decision-making pipelines to improve OOD performance in mission scenarios.
4. Hybrid symbolic-neural meta-reasoning for long-horizon planning
• Combine symbolic planning primitives with neural heuristics and learned value functions for reasoning under partial observability.
• Benchmark on hierarchical mission planning tasks with sparse rewards.

Prospective students and postdocs, please check out this page on how to get involved.

Recent Projects

• Document-Aware Vision-Language OCR with Normalized Relational Export: NASA, SMAB
• Ontology-Grounded NLP Synthesizer for Predevelopment-to-Retirement Mission Insights: NASA, SMAB
• Ensemble Prognosticator with Uncertainty-Aware GYR Decisioning for SMD Flights: NASA, SMAB
• Spatiotemporal Convolutional Prognostics for X-59 Near-Field Disturbance Detection: NASA, ASAB
• Scalable Corpus-Level Labeling and Hierarchical Folder Abstraction Engine: NASA, SACD
• Mass-Scale Multiformat Semantic Tagging and Contextual Archive Summarization: NASA, AETC
• High-Precision Invariant Mass Regression with Physics-Informed Networks: [WGU 2022]
• Zero-Shot Affective Inference via Prompt-Tuned Language Models: [WGU 2022]
• Generative Latent Modeling for Unlabeled Earth-Observation Feature Discovery: [WGU 2021]