Derek Chen Email & Phone Number

California, United States

Soleda is building a platform of AI Agents You Can Trust: - Going beyond just LLM-based co-pilots, our agents are equipped with a unique NLU module (natural language understanding) that explicit listens for and remembers user preferences - When a request or scenario is unclear, the agent intelligently asks clarification questions to resolve the ambiguity - Our agent also knows how to act appropriately due to cutting-edge reasoning capabilities trained with deep reinforcement learning (not prompting!) - This is all made possible by my prior research in scalable methods for labeled dialogue data collection

New York, Ny, Us

Data Augmentation, Few-shot Learning, Data Denoising, and Conversational AI: - Novel research on data efficiency and dialogue systems, including published work at EMNLP and NAACL - Interacting with third party vendors and internal annotation team for efficient data collection through Active Learning - Established custom data format for representing dialogues to support natural language understanding and summarization features - Oversee the data pipelines for research, applying concepts of data augmentation, learning with noisy labels, and data quality management

Palo Alto, California, Us

• Collaborated in design of end-to-end, production ready task-oriented dialogue system, including user intent tracking, policy management and response generation.• Owned the training and development of user intent classification models that served as the first piece of the pipeline described above. Responsibilities also included data collection and labeling by managing crowdsourced contractors.• Supported implementation of automatic evaluation pipeline of arbitrary dialogue models in production• Ran experiments testing SotA sentence embeddings including ELMo (AllenNLP), USE (Google), Infersent (Facebook), Distributed Sentence Representation (MILA/Maluuba), GPT vectors (OpenAI)• Taught introductory courses on building end-to-end dialogue agents for real-world settings using RL

Lafayette, Louisiana, Us

• Served as primary product owner for Google project and established daycares as a new line of business• Outlined strategic priorities through tickets, specs and wireframes in order to grow key metrics• Boosted productivity of employees by streamlining system for scheduling meetings around the office• Spent days working as a traditional product manager and nights studying CompSci foundations

Mountain View, California, Us

• Lead the development of the company's flagship product, the Nanodegree, a new credential for getting students job-ready in the modern workforce• Developed roadmap and oversaw execution of strategy to turn Nanodegrees into the main revenue driver of the company• Collaborated with content team to consistently raise engagement levels by ~10% every month• Personally pushed code to production with Angular JS framework onto pages that reached millions of users per month• Assisted marketing team in raising acquisition of paid Nanodegree students by 3x during a single quarter, with continued growth in the following months

Rockville, Md, Us

• Created a company in the EdTech industry by recruiting a team, leading product development, and iterating on releases to reach thousands of users• Enlisted half-dozen designers and developers to work on the project through sheer leadership• Programmed major portions of the app using Backbone/Underscore, JavaScript, Ruby on Rails, PostgreSQL, and Redis• Established full system of agile software development with TDD, bug tracking, and monthly release cycles• Conducted user testing, quantitative analysis, and surveys to discover insights for guiding product strategy. Good amount of carry-over knowledge gained while doing work at Apollo Education.

Redwood City, Ca, Us

• Owned the product roadmap for the company’s mobile strategy, driving 35% annual lift in visitor value• Defined, tracked, and analyzed metrics to assess product performance and inform new growth opportunities• Worked directly with engineers and designers to drive over $3 million in annual gross margin impact per quarter• Analyzed product performance across LTV, merchandising, retention, virality, conversion and cohort analysis

Portland, Or, Us

• Gained hands-on experience in Web Analytics, A/B Split-testing, Dashboard creation, and Financial Modeling• Designed and optimized $100,000+ marketing campaigns, including ownership of client relationship• Worked directly with clients to provide suggestions on prioritizing the product roadmap and marketing strategy

B.S., Business Administration

Nlp Research, Computer Science

Master Of Science - Ms, Computer Science

Foundations Of Ai, Computer Science

B.A., Mass Communications