Principal Applied Scientist, Ranking & Relevance

Machine Learning Scientist (MLS-2) – Ranking & Personalization

Role Description

As a Machine Learning Scientist (MLS-2) in Nykaa’s Search team, you will focus on the core of our discovery engine: relevance and ranking. Your mission is to ensure that every search query—from "red lipstick" to "summer skincare"—surfaces the most relevant, personalized, and high-converting products. You will build and scale Learning to Rank (LTR) models that balance user intent, historical behavior, and real-time signals to create a world-class shopping experience.

What are we looking for?

Education:

• Bachelor’s or Master’s Degree in Computer Science, Machine Learning, Information Retrieval, or a related quantitative field.

Experience:

• 3+ years of hands-on industry experience in Machine Learning.
• Proven experience in building and deploying ranking or recommendation systems at scale.
• Prior experience with Search Engines (e.G., Elasticsearch, Solr, OpenSearch) is highly preferred.

Technical Skills:

• Learning to Rank (LTR): Strong understanding of ranking objectives and algorithms (e.G., LambdaMART, RankNet, or Gradient Boosted Trees).
• Personalization: Experience in developing user-state signals and embedding-based retrieval to personalize search results.
• Signal Engineering: Proficiency in feature engineering for high-cardinality data, including CTR, conversion rates, and real-time popularity signals.
• Evaluation Metrics: Expertise in Information Retrieval metrics such as NDCG, MRR, MAP, and Precision@K.
• Tools: Proficiency in Python, SQL, and Spark;
  experience with ML platforms like SageMaker or Databricks.

Soft Skills:

• Strong analytical mindset with the ability to diagnose "why" certain results are ranked the way they are.
• Collaborative approach to work with Product and Engineering squads on A/B testing and production rollouts.

Responsibilities

• Ranking Optimization: Develop and refine Learning to Rank models to improve search relevance and conversion across Nykaa's platforms.
• Personalization Engine: Incorporate user-level preferences and historical interactions into the ranking stack to deliver a "segment-of-one" experience.
• Query Understanding: Enhance the search pipeline with models for query expansion, intent classification, and spell correction.
• Multi-Objective Ranking: Implement strategies to balance competing objectives like relevance, revenue, stock availability, and diversity in search results.
• Experimental Rigor: Design and analyze A/B tests for ranking changes, ensuring that every model deployment drives measurable improvements in North Star metrics (e.G., Search CVR, Add-to-Cart).
• Signal Productionization: Collaborate with data engineers to build robust pipelines for real-time and batch ranking signals.

What You’ll Learn & Grow Into

• Scaling ranking systems that handle millions of queries and tens of thousands of products in real-time.
• Deep mastery of modern Information Retrieval (IR) and Vector Search (Semantic Search) architectures.
• Opportunities to contribute to the next-gen AI-driven discovery roadmap, including neural ranking and GenAI integration.