
Edward Peng
AI Product Consultant for OpenAI and ChatGPT Apps
Skills

See my services

Work experience
AI Product Quality & Evaluation Lead
Xiaomi • Full-time
May 2018 - May 2025 • 7 yrs
Worked on Xiaomi’s XiaoAI voice assistant, one of Xiaomi’s large-scale AI products, and was part of its early product team. I focused on AI product quality, semantic understanding, voice interaction, wake-word evaluation, in-car voice assistant evaluation, user experience, and model output quality. I designed measurable evaluation methods, built quality metrics, analyzed failure cases, and worked with algorithm and product teams to improve AI performance. I also explored the use of GPT and prompt engineering in AI evaluation workflows. In one project, GPT-based evaluation reached 96.6% agreement with human annotation. This experience gave me strong practical knowledge in AI product management, LLM evaluation, prompt design, data analysis, and turning AI capabilities into real user-facing products.