We are seeking a Senior Machine Learning Scientist – GenAI to lead applied research and production of LLM-based products across our clients Data platform. You will work on problems social commerce ecosystems: personalized feed ranking, conversational shopping, automated content generation, and intelligent seller assistants.
You will own models end-to-end from ideation to production, with direct impact on GMV, user engagement, and creator economy metrics.
What You'll Do
- GenAI for E-commerce: Build and deploy LLM applications for core business flows: conversational product search, AI shopping assistant, auto-generated product titles/descriptions, review summarization, and livestream content understanding
- Recommendation & Ranking: Integrate LLM embeddings and reasoning into large-scale recommendation systems. Improve cold-start, cross-domain transfer, and long-tail discovery using generative models
- Multi-Modal AI: Develop models that understand video, image, text, and user behavior signals to power shoppable content, visual search, and creator tools
- Model Optimization: Research and implement SFT, RLHF/DPO, RAG, and agent frameworks tailored for e-commerce data. Optimize for latency, cost, and quality at 100M+ daily inference scale
- Evaluation & Safety: Design offline and online evaluation for GenAI features. Build guardrails for brand safety, policy compliance, and hallucination mitigation in user-facing flows
- Technical Leadership: Drive the GenAI roadmap for one or more product verticals. Mentor junior scientists and set engineering best practices for LLM development
- Cross-Functional Impact: Partner with Product, Data, Engineering, Ops, and Trust & Safety to launch experiments and measure impact on CTR, CVR, GMV, and DAU
Minimum Qualifications
- Education: MS or PhD in Computer Science, Machine Learning, AI, or related field. Research background is a strong plus
- Experience: 5+ years of industry ML experience, including 2+ years building and shipping deep learning models to production at scale
- ML Expertise: Deep knowledge of NLP, recommender systems, and deep learning: transformers, embeddings, sequential models
- Hands-on experience with LLMs: pre-training, fine-tuning, PEFT methods, prompt engineering, RAG
- Strong coding skills in Python and proficiency with PyTorch, TensorFlow, or JAX
- Systems & Data: Experience with large-scale data processing: Spark, Flink, Hive. Familiar with distributed training and high-QPS inference systems
- Product Mindset: Track record of delivering ML solutions tied to business metrics through A/B testing and iterative launches
- Domain Bonus: Experience with e-commerce, marketplace dynamics, short-video, social commerce, or ads ranking is highly preferred
