Brian Riordan Email & Phone Number

New York City Metropolitan Area

Applied science and ML engineering management; hands-on generative AI R&D and production ML; generative AI strategy

Princeton, Nj, Us

Main focus: Led ETS R&D on deep learning and LLMs for natural language understanding (NLU) in education.Led R&D and deployments to production (both hands-on development and project leadership): - Applications of LLMs: Early-stage R&D on efficient fine-tuning, instruction tuning and reinforcement learning, retrieval-augmented LLMs (RAG), reasoning, applying new LLM architectures - Large-scale NLP deployments: Algorithms for content assessment leveraged in multiple large-scale deployments (large state summative assessments, high-takes language proficiency assessments) - Real-world impacts: Developed idea detection algorithms deployed in an intelligent tutoring app in California schools Developed high performance computing (HPC) infrastructure for ML workflows: - Led the first large-scale deployment of deep learning-based models at ETS - HPC issues in distributed training, hyperparam optimization, and real-time inference, with on-prem/cloud computeR&D team leadership - Led agile teams for automated scoring and NLU for edtech applications - Served as PI or Co-PI on multiple NSF and IES projects; secured nearly $1 million in external grants for NLP R&D for STEM education (collaboration with UC-Berkeley) - Mentored junior researchers, engineers, and interns - Cultivated R&D collaborations with academic institutionsR&D focus areas: - NLP: large language models (LLMs), short answer scoring, essay scoring, chatbots / conversational AI / dialog, entailment/NLI, information extraction, knowledge graphs, paraphrasing, question answering, sequence labeling / token classification, text classification - ML: efficient transformers, human-machine interaction, explainability/interpretability, learning with limited data (zero/few-shot learning, meta-learning, transfer learning, weak supervision), multi-task learning, representation learning (contrastive learning), robustness

Princeton, Nj, Us

Main foci:- Carried out deep learning R&D for edtech applications including: short answer scoring, question generation- Served as lead scientist for ETS c-rater content scoring system- Set a SOTA on the ASAP-SAS short answer scoring datasetR&D focus areas: - NLP: short answer scoring, essay scoring, information extraction, sequence labeling / token classification, text classification, text similarity - ML: deep learning (transformers), explainability/interpretability, learning with limited data (few/zero-shot learning, meta-learning, transfer learning), multi-task learning, robustness

Woburn, Ma, Us

Main focus: Machine learning and NLP in diverse defense and intelligence applications. Led numerous R&D projects, carrying out both hands-on development and project leadership: - Secured more than $8 million in external research funding; served as Principal Investigator or lead algorithm developer on all projects - Clients included Defense Advanced Research Projects Agency (DARPA), Office of Naval Research (ONR), Air Force Research Laboratory (AFRL), Army Research Laboratory (ARL), Office of the Secretary of Defense (OSD), Defence Research and Development Canada (DRDC) - R&D for natural language understanding applications - R&D with Bayesian machine learningProject management and R&D team leadership: - Product lead, text analytics - Led agile R&D project teams for diverse applications - Co-led Aptima's CMMI-based agile transformation - achieved CMMI Level 2 certification in an agile project management framework and began work on Level 3 certificationCompany awards: - Best Technical Achievement, Best Paper, Best Initiative, Best Teamwork - First staff member to win all four company awardsR&D focus areas: - NLP: topic modeling, authorship attribution, debate stance classification, entity resolution, natural language inference (NLI) / textual entailment, question answering, question generation, sentiment analysis - ML: Bayesian statistics, decision making under uncertainty (especially Markov Decision Processes, inverse reinforcement learning), kernel methods (Gaussian Processes), inference algorithms (MCMC, variational), latent factor models (especially topic models, non-parametric Bayesian models), probabilistic graphical models

Bloomington, Indiana, Us

R&D focus areas: - NLP: representation learning for NLP - ML: Bayesian statistics, generalized linear mixed models (GLMMs), latent factor models (topic models)

Jp

Expert-level Japanese language proficiencyJapanese-English translation and interpretation

Ph.D., Computational Linguistics, Cognitive Science

B.A., Linguistics, Anthropology, East Asian Studies

M.A., Computational Linguistics

Education record