Data Scientist III

Job Role: Data Scientist 3 – Ads & Monetization / Pricing Optimization

Experience Required: 5–7 years

Location: Bangalore | Karnataka

About Swiggy And The Team

Swiggy is Indias leading on-demand delivery platform, leveraging data science and cutting-edge AI to redefine convenience for millions of customers. The Food-DS team works at the intersection of machine learning, advanced architecture, and applied research to shape AI-first systems that directly impact customer experience and business growth. The team values cross-functional collaboration, open sharing of ideas, and continuous innovation to roll out machine learning and AI solutions at scale.

About The Role

As a Data Scientist 3 (DS3), you will act as a Technical Stream Lead within the Ads, Pricing, and Discount Optimization sub-domain. You will take end-to-end ownership of scoping, designing, and delivering complex algorithmic decision engines (typically 2–3 related models or experiments over a quarter). You will operate with high autonomy—translating open-ended monetization and incentive allocation challenges into mathematical and ML frameworks, balancing ROI trade-offs, and driving unit economics (L0/L1/L2 metrics) while mentoring junior data scientists.

What Qualities Are We Looking For

Technical Expertise & Problem Formulation

• Experience & Ownership: 4–6 years of experience building and deploying scalable decisioning engines, causal models, or algorithmic ad-tech architectures in production.
• Problem Formulation: Proven ability to break down open-ended pricing, spend allocation, or ad bidding problems into DS/ML sub-problems, defining appropriate targets, constraints, and objective functions.
• Reinforcement Learning & Decision Systems: Hands-on experience with Contextual Bandits, Multi-Armed Bandits (MAB), Markov Decision Processes (MDPs), Q-Learning, Policy Optimization, and Offline Reinforcement Learning for dynamic ad allocation and coupon targeting.
• Causal Inference & Economics: Deep knowledge of Uplift Modeling, Counterfactual Evaluation, Price Elasticity Modeling, and Causal Inference to measure incremental business lift and avoid deadweight loss.
• Optimization & Decision Engines: Expertise in Constrained Optimization, Mixed-Integer Programming (MIP), Bayesian Optimization, and Budget Allocation frameworks to design production-grade Decision Engines.
• Engineering Excellence: Proficient in distributed processing and execution (PySpark, Python, PyTorch/TensorFlow, CVXPY/Gurobi/SciPy) for real-time and batch optimization systems.

• Technical Stream Leadership: Ability to act as the Single Point of Contact (SPOC) for Ads/Discount optimization streams, guiding DS1/DS2 engineers, managing delivery coherence, and setting stakeholder expectations.
• Domain Metric Ownership: Ability to connect DS levers (e.g., ad CTR, eCPM, take-rate, ROI, gross margin) to core business outcomes, root-causing metric shifts (L2 to L1), and proactively identifying spend efficiency opportunities.
• Operational Excellence: Rigor in maintaining system hygiene owning root-cause analyses (RCAs), minimizing tech debt, optimizing pipeline compute costs, and ensuring auction/pricing stability during peak traffic events (e.g., festivals, IPL).
• Pragmatic Innovation: Capability to adapt state-of-the-art research (e.g., novel bandit architectures, causal uplift methods, offline RL evaluation) pragmatically to Swiggys scale and operational constraints.

• Lead Project Streams: Scope, design, and implement solutions for dynamic pricing, ad ranking/bidding, and discount allocation; write comprehensive approach notes, design docs, and phased experiment plans
• Build Algorithmic Decision Engines: Architect optimization engines combining Contextual Bandits, Causal Uplift Models, and Constrained Optimization to allocate ad slots and promo budgets at scale.
• Drive Metric Accountabilities: Function as tech lead for a sub-domain, owning at least 1 Key Result (KR) and using data insights to justify model decisions to Monetization, Product, and Business teams.
• Elevate Engineering Standards: Implement best practices within your pod—including code reviews, test-driven pipelines, version control, on-call hygiene, and async-first technical documentation.
• Mentor & Collaborate: Actively guide and pair with DS1s and DS2s, streamline cross-functional communication, and contribute to pod-level roadmap discussions and technical publications/blogs.

• Impact at Scale: Work on high-velocity systems driving real-time decisions for millions of orders.
• Technical Autonomy: Lead sub-domains end-to-end with the space to introduce state-of-the-art techniques.
• Collaborative Growth: Work in a high-density learning environment alongside cross-functional experts in Engineering, Product, and Business.