Machine Learning Engineer

About Us

We are an AI-powered local discovery platform that helps people explore cities and find great places effortlessly. By combining conversational AI, interactive maps, and community-driven content, we enable users to discover restaurants, attractions, and unique local experiences through natural, intuitive interactions. We are building the next generation of AI-powered discovery, reimagining how people explore the world around them.

About the Role

We're hiring ML engineers to build the core intelligence behind our search, recommendation, and multimodal retrieval systems.

This is a high-impact, hands-on role for engineers who want to apply machine learning, deep learning, and generative AI to real product problems at scale. You'll work on retrieval, ranking, relevance, content understanding, embeddings, and multimodal models, and help turn research ideas into production systems that improve user experience and business metrics.

Responsibilities

• Build and deploy ML models for search, recommendation, and discovery.
• Improve multimodal retrieval across text, image, video, and structured data.
• Develop and fine-tune LLMs and vision-language models for production use cases.
• Work on relevance problems including query understanding, recall, ranking, embeddings, and RAG.
• Run experiments, analyze results, and drive measurable improvements in KPIs.
• Process large-scale data and build practical, scalable ML systems.

Requirements

• Strong experience in machine learning applied to search, recommendation, ads, feed ranking, or relevance.
• Strong fundamentals in ML, deep learning, NLP, or multimodal learning.
• Hands-on experience with retrieval, ranking, embeddings, relevance, optimization, or RAG.
• Experience with multimodal models is strongly preferred.
• Familiarity with LLM fine-tuning methods such as SFT, LoRA, PEFT, RLHF, or DPO is a plus.
• Strong programming skills in Python, Go, or C++.
• Comfortable working in a fast-moving, high-ownership startup environment.
• Fluent in Chinese is a must.