Principal ML Scientist - Applied Sciences

Principal Machine Learning Scientist - Applied Sciences(New Initiatives)

About Nykaa

Nykaa is India's leading beauty and lifestyle destination. We are a consumer-tech company, offering a portfolio of beauty, personal care, and fashion products through our online platforms and retail stores.

At our core, we are powered by technology and data, and our Data Science team is at the forefront of creating intelligent, personalized, and seamless experiences for millions of customers. As we enter the next phase of growth, we’re investing deeply in foundational ML systems that will enable real-time decisioning, large-scale automation, and GenAI-powered experiences.

Role Overview

As a Principal Applied Scientist, you will be a senior technical and thought leader responsible for driving the scientific roadmap for our core business operations. You will tackle our most complex and ambiguous challenges in

Supply Chain Optimization, Demand Forecasting, Marketing Personalization, and Fraud Detection .

This is a deeply technical, hands-on role focused on developing and deploying robust, scalable solutions that drive tangible business outcomes. You will leverage your expertise in machine learning, statistics, and optimization to build models that directly impact inventory efficiency, marketing ROI, and platform integrity. This role requires a blend of deep scientific expertise, strong business acumen, and a passion for mentoring and elevating the entire data science community at Nykaa.

Key Responsibilities

Identify and frame high-impact problems across supply chain, marketing, and platform integrity, translating business ambiguity into clearly scoped scientific programs with measurable ROI.

Develop advanced demand forecasting models using statistical and machine learning methods such as Prophet, ARIMA, LSTMs, and transformer-based time-series architectures to predict demand and returns at multiple hierarchies (SKU, region, season, channel).

Optimize supply chain and fulfillment networks through data-driven algorithms for warehouse placement, SKU allocation and inventory planning etc, leveraging genetic algorithms, mixed-integer programming, and reinforcement learning.

Enhance delivery predictability by modeling EDD (Estimated Delivery Date) using spatiotemporal and supply-side signals (carrier capacity, warehouse load, regional demand patterns).

Drive marketing science initiatives by developing models for coupon optimization, price elasticity, uplift modeling, and marketing attribution, improving campaign efficiency and user-level personalization.

Detect and prevent fraud and abuse through graph-based anomaly detection, temporal embeddings, and unsupervised outlier detection — protecting against fake reviews, referral abuse, and promotion misuse.

Model and minimize product returns, combining behavioral data, text / image feedback, and fulfillment patterns to proactively identify high-return-risk SKUs and customer cohorts.

Build decision-intelligence systems that integrate forecasting, supply, and marketing signals to enable real-time business decisioning and scenario planning.

Lead cross-functional initiatives, mentoring scientists and engineers, guiding technical reviews, and ensuring scientific work translates into robust, production-grade solutions.

Collaborate closely with Product, Engineering, Operations, and Marketing leaders to ensure models are integrated seamlessly and deliver sustained business impact.

Qualifications & Skills

Experience : 10+ years of experience in Applied ML, Operations Research, or Quantitative Modeling, with proven delivery of large-scale ML systems in production.

Education : PhD or Master’s in a quantitative discipline (Operations Research, Statistics, Econometrics, Computer Science, Applied Mathematics, or related).

Expertise in Time-Series Forecasting,

including classical

(ARIMA, ETS) and modern (LSTM, Transformer, Temporal Fusion Transformer, DeepAR) methods.

Experience in Optimization & OR techniques — Linear / Integer Programming, Genetic Algorithms, Reinforcement Learning, and heuristic optimization.

Strong understanding of Causal Inference, Bayesian Modeling, and Probabilistic Forecasting.

Hands-on proficiency in Python, PyTorch, TensorFlow, XGBoost, Prophet, Scikit-learn, and PyMC / Stan.

Comfort working in large-scale data environments such as Databricks, AWS SageMaker, Spark, and MLflow for model tracking and deployment.

Ability to translate data science into tangible business outcomes — cost reduction, margin uplift, improved delivery SLAs, reduced fraud losses.

Strong storytelling and communication skills to influence senior stakeholders and non-technical partners.

Prior experience in e-commerce, retail, logistics, or marketing analytics is highly desirable.