Yinghong L.

* Set multi-year technical vision with real business and product impact based on LLM, Speech, Multimodal and Causal Inference ML. * Grew the team to 3x size within one year through interviewing, onboarding and coaching Deep Learning, Causal Inference and Engineering talents. * Guided and supported the team in launching highly impactful algo products and pioneering new algo services. * Strengthened and established strategic partnerships and collaborations with backend, platform and product teams.

Sample projects: * Developed Multimodal ML models for evaluating audio-visual quality of localization. * Developed and productized Causal Inference models for optimizing investment in localization.

Sample projects @ Ad Market Dynamics team: • Led an end-to-end project for improving the ads auction algo design, from benchmarking different algos via simulations to driving the measurement and productization steps. • Designed and analyzed several rounds of geo experiments to test the effectiveness of bidding algorithms using Causal Inference.

• Constructed pipeline to simulate evolutionary dynamics under different conditions. • Analyzed synthetic data to develop theoretical model on robustness of evolutionary landscapes. • Evaluated statistical power of different experimental paradigms.

· Explored different modeling options of time to event data to optimize acquisition. · Analyzed multilevel observation data with Bayesian Hierarchical models to answer critical product questions.

• Designed and constructed genetic circuses for estimating the inhibition efficiency of zinc finger domain, a widely used transcription factor backbone.

• Led the team USTC-China of 16 undergraduates to design and accomplish the project at iGEM 2011. • Won Gold Medal. Advanced to Championship.