Gregory Toepperwein Email & Phone Number

● Provide tactical direction and leadership for language translation and NLP initiatives, overseeing amultidisciplinary team to ensure alignment with company goals and objectives.● Continuously adapt machine learning (ML) translation tools and seek methodological improvementsfor low-resource language translation, leveraging innovative approaches and techniques.● Champion innovation, introducing improvements at all stages of the translation process andcoordinating their implementation across organizational boundaries.● Structure the overall translation flow, integrating human and machine translation efforts. Hold a deepunderstanding of process beyond its immediate technical components. Contribute to defining userroles, feedback mechanisms, and model baseline creation to optimize translation quality.● Lead and coordinate the efforts of a diverse team of NLP engineers, data scientists, andcomputational linguists, fostering collaboration and driving discovery.● Contribute to corporate and public understanding of our language translation and NLP capabilitiesand how they support company objectives.● Develop and implement innovative strategies to enhance translation quality and drive technicalexcellence across all NLP-related projects, including the definition and evaluation of metrics.● Organize and document team efforts.

• Coordinated NLP team efforts across projects for language translation, medical conversation summarization, symptom detection, and text evaluation in foreign languages • Continuously adapting ML baseline translations and seeking methodological improvements for low-resource language translation• Crafted model prompt for LLM summarization of medical conversations, used both in product and for generation of training set for future models• Directed manual NLP dataset creation effort of ~10 people for detailed markup of a small corpus; effort included scheme definition, motivating contributes, daily tracking, addressing issues, and quality review• Developed definition management scheme to propagate English concepts to arbitrary target language allowing for enforcement of translation consistency• Organized review scheme to get rapid translator feedback on proposed AI translationsCrafted metrics for evaluating translation naturalness in low-resource languages• Refined symptom synonym definitions in support of medical app identifying patient services• Experimented on training text selection to get proper balance of thematic influences and vocabulary coverage with minimal input• Setup documentation and tracking around identified entities and concepts sensitive to proper translation• Guided selection of which low-resource languages to pursue based on ability to leverage existing translation infrastructure• Identified dialect differences from translators without personal knowledge of languages• Scripted data cleanup in support source text parsing correction, filtering of reference documents, and model ingestion

Portland, Oregon Area

• Guided students through entire pipeline of modeling activity, imparting deep knowledge ofstatistical and machine learning algorithms to next generation of data scientists• Built rapport, listened, and created a supportive environment for clients in which to deliver professional tutoring services most attuned to their individual needs• Employed Natural Language Processing (NLP), using deep-learning architectures to solveproblems including text classification, image captioning, and sentiment analysis• Communicated complex findings and design concepts to clients of all technical abilities• Consulted on machine learning projects, helping to design, plan, implement, and automatecomplex workflows for medical event detection, recommender systems, and video analysis• Partnered with clients on data analysis problems, demonstrating ETL techniques andimparting experience with data organization, infrastructure, and strategy• Developed algorithms hand-in-hand with novice students, translating stakeholder requestsinto sound methodology• Provided technical leadership to advanced students and professionals to enable self-sufficiency and long-term independence• Built AI and ML models, creating end-to-end (E2E) solutions from inception to deployment• Designed and produced training course (video, slides, and homework) for corporate clienttransitioning team to Python and Pandas• Engaged in ongoing brand development and marketing of services

Hillsboro, Or

• Built, implemented, and iterated on data-focused models for photoresist development andsubsequent processing to enable advanced optical pattern correction (OPC) on reticles usedin Intels 14nm and 10nm technology nodes• Partnered with computational and lithography stakeholders to understand and probebusiness needs, providing on-demand support to divergent activities that enabled teammembers, fostered innovation, and ensured all projects advanced in accordance withcorporate schedule• Deployed cutting-edge solutions to production-level systems (see Patents section)• Worked in newly-formed, small, dynamic data-science team with an open-ended mandate toimprove modeling methodology, metrics, and computational tools for novel reduction of model error through analysis of simulated images; primary resource for training of larger OPC team on said techniques • Rapidly developed high-impact photoresist models in support of high-visibility products• Effective partner to product managers, working in cross-functional teams to continuallyanalyze yield-limiting liabilities and provided process insights to drive business decisions• Contributed to highly-collaborative environment supporting junior members through codereviews, documentation, group discussions, and individual training• Relentlessly pursued quality, executing deep dives into to identify and diagnose datainconsistencies and errors, document data assumptions, and foraged to fill data gaps• Led efforts to develop and improve Python programming acumen across entire team; included use of Pandas, Machine-Learning toolkits, and derived in-house tools.• Produced and advanced internal code to improve task automation in a linux environment

• Investigated effect of nanoparticle inclusions on the structural properties of the underlying polymer matrix via coarse-grained models• Engaged in collaboration with Intel and Cornell University to develop commercially feasible processes using traditional resists with super-critical CO2 as an environmentally benign solvent using various additives (quaternary ammonium salts, reactive additives, etc.) to enhance solubility• Explored effect of nanoparticle size, aspect ratio, concentration, and surface chemistry on mechanical strength of polymer nanocomposites• Modeled ionic liquids to determine mechanical properties and feasibility as resist developer• Studied void formation in nanocomposites to understand the role inclusions play in failure• Mentored undergrad research assistant for three semesters, guiding student’s development of programming skills and application of advanced simulation algorithms• Developed new molecular model for Siloxane-based alternative lithographic solvents• Investigated polymer dynamics under confinement, including free surface relaxation andnanocomposite creep response• Modeled molecular glass photoresists in multiple solvents• Maintained 4 servers with 100+ nodes for use by all members of a large simulation-basedresearch group, responsible for interfacing with outside resources, recovering from systemfailures, and day-to-day operations (quota management, bug fixes, etc.)• Managed resource networking with Condor software as part of the Grid Laboratory ofWisconsin (GLOW)

• Contributed to semiconductor lithography projects by conducting experiments, completing research, and performing engineering calculations• Performed Factory Acceptance Testing for 193nm interference lithography (IL) tool from newsupplier as part of Sematech’s R&D pilot line• Prepared test wafers on photoresist track including: spin-coating polymer onto wafers, pre-baking, post-baking, developing, and rinsing• Used VUV ellipsometer to test film thickness and calibrate spin speed for testing wafers• Acted as primary operater for new IL tool• Produced dose-focus arrays on new IL tool and manually evaluated results on SEM to characterize process window• Interacted with IL tool supplier to address issues with objective crashing into wafer• Modeled the impact of polarization and film stack parameters on 45nm photoresist images usingProlith simulation software• Collected and analyzed data for acceptance testing of reticle and prepared abbreviated summary of reticle for divison-wide reference

Doctor Of Philosophy (Ph.D.), Chemical Engineering

Bachelor Of Science (Bs), Chemical Engineering