Asurion is seeking a Principal Data Scientist to join our Data Science team. An individual in this role will be designing, developing, implementing, and fine-tuning Large Language Models to build one-of-a-kind autonomous sales agents in the context of tech support. The candidate will work on the cutting-edge of LLMs as we aim to leverage them for sales where the tasks are better defined and narrow, but margin of error is very low. Leveraging AI to be offer a highly targeted and personalized sales experience while setting state-of-the-art standards for compliance and hallucinations. They should be well versed on modern massive neural language model architectures, including BERT-like and GPT/autoregressive-like LLM's, neural language generation, AI agents, few-shot multitask learning etc. The candidate should also know the limitations of LLMs and how to design solutions and services keeping those in mind. Gather and analyze large volumes of unstructured, text data, evaluates scenarios to make predictions on future outcomes and supports decision making. An ideal candidate would have excellent communication (written and oral) and presentation skills, including creating and sharing complex ideas to peers, cross-functional partners and stakeholders. The candidate will also offer technical mentorship and guidance to junior data scientists.
Role and Responsibilities:
- Utilize Natural Language Processing to analyze speech and text data, build chatbot and expert systems
- Implement existing Large Language Models such as GPT, LLaMA, etc
- Refine Large Language Models via prompt engineering, Agentic frameworks, fine-tuning, reinforcement learning, or transfer learning
- Lead and mentor a team of scientists, driving high-impact technical initiatives with broad scope and visibility
- Discover and formulate business problems and create data science solutions
- Develop prototypes for new data product ideas
- Design and deploy complex, optimized and large-scale machine learning algorithms and generative models to improve Asurion’s wireless, connected home, and appliance customers’ experiences
- Analyze and interpret the results of product experiments
- Work closely with product managers to identify and answer important product questions that help improve outcomes
- Communicate findings to product managers and development groups
- Drive the collection of new data and the refinement of existing data
- Regularly invents new and novel approaches to problems; takes initiative and breaks down barriers to solve problems; recognized within team as the source of solutions
Qualifications:
- Master's or PhD preferably in Machine Learning role with at least 4+ years hands-on ML experience
- 2+ years of experience leading a team of scientists or engineers, with a strong track record of recruiting and/or mentoring junior team members
- Specialized practical experience in Generative AI
- Expert in Natural Language Processing tasks
- Applied experience in deploying NLP based solutions, including Gen AI, at scale in a cloud environment like AWS, GCP or Azure.
- Prior experience in applying machine learning for customer experience use cases
- Comfortable manipulating and analyzing complex, high-volume, high dimensionality data from multiple sources
- Comfortable in using advanced statistical data modeling techniques and tools
- Experience using Python and working with large scale systems
- Knowledge of data processing on Hadoop programming environments (e.g. Spark/Hive)
- Familiar with optimization techniques and underlying machine learning algorithms concepts
- Solid statistical knowledge; familiarity with hypothesis testing, experimental design and time series
- Familiarity with Linux/Unix/Shell environments
- Self-driven and able to ask and tackle the most important analytical questions with a view on driving product impact
- A strong passion for empirical research and for answering hard questions with data
- Result-driven and focused self-starters, great communicators that follow through on every initiative
- Love the responsibility of being individually empowered
- Ability to communicate complex quantitative analysis in a clear, precise, and actionable manner