• Leading design of conversational AI experiences, from flows to prototypes

• Iterating on prompts and UX, handing finalized specs to engineering

• Set up user testing and incorporate results into conversation flows

• Present designs to senior leadership, shaping key product decisions

• Lead wireframing and prototyping of conversational flows, turning complex user needs into clear, scalable AI interactions

• Drive end-to-end product design for Fargo, ensuring conversations are intuitive, natural, and aligned with business goals

• Present designs to key stakeholders (devs, legal, compliance) and lead User Acceptance Testing (UAT)

• Source, generate, and annotate data for intents, entities, and conversational context

• Review production log conversations and optimize model performance

• Monitor model metrics (e.g., F-scores, accuracy) and recommend retraining strategies

• Collaborate with teams to refine training data and improve overall NLP accuracy

• Led end-to-end conversational experience for an AI-powered virtual assistant on over 20 projects with Fortune 500 companies

• Partnered with client-side stakeholders to draw insights from business processes and design workflows

• Developed, annotated, and tested large training datasets using machine learning models

• Designed and managed crowdsourcing projects

• Collaborated with the R&D team on improving intent and entity recognition, and humanization of VA's responses

• Managed QA and UAT tasks, and reported conversational accuracy metrics

Designed and implemented natural language solutions to help users understand their financial insights more clearly and intuitively.

• Created rules and generated content for an automated natural language generation engine

• Integrated insights data into code

• Managed multiple sprints, coordinating dev team to deliver rapid improvements

• Optimized language outputs to improve clarity and engagement for end users

Immediately after graduating, I joined Google to help build Text-to-Speech (TTS) and Automatic Speech Recognition (ASR) systems for Ukrainian. These systems and the voice my team helped select and shape was deployed across Google products like Google Maps, Translate, etc, reaching millions of users globally.

• Developed phonotactic rules, lexicon, grammar and annotation conventions

• Transcribed over 20,000 lexicon entries using in-house tools

• Labeled text for disambiguation, expansion, and text normalization

• Provided expertise in accent analysis and selection of a voice talent

• QA-ed voice recordings and phonemic transcriptions

• Created and maintained an extensive lexical ontology

• Annotated and classified search queries

• Developed a dictionary of morphological paradigms with POS tags

• Provided analysis of linguistic evidence in criminal cases with death row verdicts

• Assisted legal counsel in court proceedings

As part of my Provost Scholarship, I was offered this position, and later invited by the Program Director to co-author one of the first comprehensive forensic linguistics bibliographies with Oxford University Press.

Key responsibilities:

• Managed program correspondence and inquiries

• Maintained program's web page

• Co-edited an extensive forensic linguistic bibliography