Elizabeth Hutton Email & Phone Number

Los Angeles, Ca, Us

San Jose, Ca, Us

As a senior MLE on Cisco's new Collaboration AI team, I am leading the development of new NLP and speech-related AI features across Cisco's range of collaboration products including Webex Meetings, Calling, Contact Center, and Vidcast. Key accomplishments: • Took Cisco’s first ever LLM powered features from R&D to production in less than 6 months• Developed in-house models for effective topic-based segmentation of long transcripts, currently powering all downstream LLM features• Identified multiple alternative models which are cheaper but just as effective for our use cases (over $6 million/year savings) • Implemented SOTA evaluation practices and guardrails to measure LLM powered features• multiple patent applications and publications My Responsibilities Include: • SOTA model development, evaluation, and serving• prompt engineering• design, build, and test our API's• conduct ML experiments and communicate results• gatekeep all things data, including curation and annotation• lead time/money saving data generation, validation, and augmentation initiatives • represent the AI team at talks and customer meetings

San Jose, Ca, Us

As the lead MLE on the Webex Contact Center AI Team, I took the business's first-ever in-house AI solutions from R&D to production. Accomplishments:• 3 patents pending for development of novel technology • productionized WxCC's first-ever native AI solution, Cisco Agent Answers, a retrieval augmented generation (RAG), knowledge management, and question answering system • developed and trained an LLM to summarize conversations for automatic call wrap-up • enabled the team's rapid iteration with the creation of an easy-to-use evaluation platform for tracking ML metrics, versioning data/models, and communicating results through multi-media reports

San Francisco, Ca, Us

- Built text2topics, a topic modeling package that uses NLP to automatically extract the key concepts from unstructured documents.- Collaborated with ML engineers from Ping, a company that provides automated timekeeping software.- Used AWS for model training and Git for version control, with pull requests reviewed by a senior AI professional.

Los Angeles, Ca, Us

- ­Led a research project on statistical learning of auditory stimuli. -­ Developed and tested neural network models in TensorFlow for experiment simulation.- Created experiments in MATLAB and Paradigm. Performed analyses in R. -­ Analyzed behavioral data from 8 experiments on 350+ subjects. ­- Managed and trained a team of 6 research assistants.-­ Co-­first authored a research paper in press at the Journal of Cognitive Science- Presented a poster at the 2017 CogSci Conference in London

Barcelona, Barcelona, Es

- Collaborated with Dr. Gustavo Deco’s computational neuroscience lab. - Modeled patterns of brain activity using fMRI data from epilepsy patients.- Adapted model’s code (Python and MATLAB) to new data set, tuned parameters, investigated and presented results. Interned over the summer at a computational neuroscience group in Barcelona, Spain under the mentorship of Dr. Gustavo Deco. I compared fMRI data of epilepsy patients with those of healthy controls using a computational model that extracts effective connectivity between regions throughout the whole brain.

Pasadena, California, Us

Duties included answering phone calls, filling prescriptions, managing billing, greeting patients.

Pasadena, Ca, Us

- Worked in Prof. Sossina Haile’s Applied Physics and Materials Science Laboratory studying the chemical properties of doped materials for use as catalysts in thermochemical water-splitting.- Performed materials synthesis, x-ray diffraction, and data analysis in Excel. - Scaled production of doped ceria for use in full-size hydrogen fuel cell prototype. - Earned a full-time paid position to continue my project over the summer.

Bachelor Of Science - Bs, Applied And Computational Mathematics

Bachelor Of Arts - Ba, Cognitive Science

Physics