Emma Ehrhardt Email & Phone Number

Indianapolis, In, Us

Leading a mature team of Computational Linguists, Language Data Engineers, and Data Annotation Specialists. We provide linguistic consulting, connect with native speaker expertise, curate language data for benchmarking and model training, and design and execute "AI Smarts" benchmarks across the full range of Conversational AI features and languages Genesys supports.

Indianapolis, Indiana, United States

I work in Genesys' Applied Research Team in Artificial Intelligence. The team I'm on bridges the gap between academia and industry by turning cutting edge academic research on NLP & ML into usable technology that can be embedded in Genesys Conversational AI products.My primary focus these days (2020-2021) is around:- Quality Evaluation of Language-related technologies. Includes scoping and overseeing projects, curating DATA (build/collect/find, annotate), designing and conducting EXPERIMENTS to assess performance of models/engines/datasets/etc, building out TOOLS & PROCESSES to support the research in a way that is standardized, efficient, & reproducible, and synthesizing the KNOWLEDGE that results from our research, spanning everything from ConvoAI Products (Bots, Agent Assist) to component language technologies (speech recognition, NER, sentiment analysis, etc).- Linguist Subject Matter Expert for ConvoAI products and features, wrt overall UX and features involving language. E.g. best practices for designing & tuning chatbot NLU models, guidance on designing test sets to verify functionality and detect side effects of new language-related product features, localization gotcha's, etc.- Team lead for a group working on the above.- I'm passionate about advocating for rigor in evaluating AI feature quality, that it's not enough to verify a few happy paths, but that rigorous evaluation is critical, across a range of dimensions as well as special attention to likely weak areas. I aim to demonstrate this in the work I do, and to encourage others to "get the big picture" that 1) evaluation is required and 2) it cannot be a trivial afterthought.

Indianapolis, Indiana Area

Indianapolis, Indiana Area

Co-led a team that owned in-house developed speech technologies spanning Call Analysis, Speech Recognition in free conversation and in grammar-driven dialogs, Speech Synthesis, and Echo Cancellation. Some of my many hats included:- Project coordinator/owner. Making sure everyone on a project gets (and stays!) on the same page, & help keep things moving forward by keeping the focus on our end goals.- Technical & Research guidance for the team members I lead directly. Challenging the team to take on increasing ownership of all areas, not just where they've worked before and are comfortable.- Liaison/Advocate. Coordinating priorities with PMs, building relationships with other Dev, Testing, Support, etc, and educating about our products with Sales, Education, etc.- Release manager / QA. Ensuring our team's releases and improvements are well-tested, fully documented, hit the appropriate release deadlines, & communicated appropriately.

Indianapolis, Indiana Area

Indianapolis, Indiana Area

- Build language resources for ASR & TTS systems (pronunciation dictionaries, text normalization rules, spelling-to-pronunciation, etc.) Multiple languages.- Design, select, collect, & evaluate speech corpora. (E.g. TTS: corpus selection driven by phonological, lexical, & prosodic concerns; voice talent coaching; ASR: Evaluate speaker demographics, linguistic coverage, acoustic environment variability, annotation quality; coordinate data collection projects with native speakers).- Develop tools to facilitate language resource development (corpus selection algorithms, dictionary error & anomaly checker, library to manipulate TTS training data formats).- Analyze speech data in order to evaluate ASR system performance, and suggest areas for future R&D work.- Identify, document, and improve team processes & best practices.- Coordinate various speech projects, from data collection & annotation efforts, to the status of ASR & TTS languages under development.

Tucson, Arizona Area

Tucson, Arizona Area

Master Of Science (Ms), Human Language Technology, 4.0

Degree could alternately be called Computational Linguistics or Natural Language Processing. Combining Advanced Linguistics, Computational.

Bachelor Of Arts (Ba), Biblical Exposition

Bible & Theology. Greek & Hebrew. Teaching methods. Learning to think deeply and act intentionally.

Linguistics

Attended UND's Summer Institute of Linguistics, exploring Linguistics as a complement to my Bachelor's degree.