Jan 2022 – Jan 2025 · 3 yrs

• Directed the full life cycle of building Conversational Agent Platform from ideation to deployment with customers.
• Reduced inference latency by 70% by moving from api-driven services to a self-hosted environment.
• Conducted user research and competitive analysis to refine conversational AI features, prioritizing impactful updates to improve user engagement.
• Designed and maintained a healthy product backlog for AI feature development, using JIRA to align sprints with stakeholder priorities.
• Defined and refined detailed user stories with clear acceptance criteria to guide engineering and ensure the delivery of impactful conversational AI features.
• Collaborated with the engineering team to fine-tune conversational AI LLM models (OpenAI GPT-4o, Anthropic Haiku, Llama) with a focus on user-centric style control and guardrailing.
• Guided engineering team on how to use open-source AI projects(PipeCat/Vocode/LiteLLM/LangChain) and other non-open-source technologies to accelerate product development.
• Helped build the evaluation process using OpenPipe to ensure that responses from RAG (Retrieval-Augmented Generation) met the bar for deployment to production.

Jan 2017 – Jan 2022 · 5 yrs · San Francisco Bay Area

• Led the global rollout of multilingual support, launching in 120+ countries to improve user engagement.
• Worked with the Data Science Team to ensure that BERT (transformer-based) search, Apache Airflow, and Knowledge Graph (Neo4J) pipelines produced verifiable results for users.
• Built out unique business intelligence stack, including real-time availability for supply vs. demand, heat mapping, and time-series analysis