Senior Data Scientist
Swiggy · Bangalore · 6–9 yrs experience · Posted 2026-06-10
Tech stack: Python, Machine Learning, Deep Learning, Generative AI, PySpark, TensorFlow, PyTorch, Vector DB, LangGraph, CrewAI, AutoGen, Computer Vision, Recommendation Systems
About the role
Official Swiggy role: Senior Data Scientist.
Official business unit: Technology.
Location: Bangalore, India.
As a Senior Data Scientist, you will lead the design and implementation of Swiggy's core Machine Learning (ML), Deep Learning (DL), and next-generation AI systems. This role requires deep-rooted expertise in classical ML, DL, Large-scale Search, and Recommendation Systems. You will act as a hands-on technical leader, balancing the deployment of cutting-edge systems with influencing business KPIs to enhance the customer experience.
Architect Hybrid Systems: Build systems that leverage both Generative AI (Agents/LLMs) and traditional ML (Ranking/Search) to solve complex search and discovery problems.
Scale High-Performance Models: Design and train large-scale Machine Learning and Deep Learning models for Swiggy's recommendation and ranking pipelines.
Responsibilities:
- Architect Hybrid Systems: Build systems that leverage both Generative AI (Agents/LLMs) and traditional ML (Ranking/Search) to solve complex search and discovery problems.
- Scale High-Performance Models: Design and train large-scale Machine Learning and Deep Learning models for Swiggy's recommendation and ranking pipelines.
- Innovation & Benchmarking: Stay ahead of the curve by benchmarking emerging trends and contributing to the global AI community through research or blog publications.
- Cross-functional Leadership: Lead end-to-end projects from problem statements to production, ensuring scalable and cost-effective AI deployments.
Qualifications:
- Experience: 6-9 years in AI/Data Science, with a proven track record of deploying models in production.
- NLP: Hands-on experience with transformer-based open-source SLMs and fine-tuningEncoder models (e.g., Nomic, RoBERTa, GTE etc.,) for specialized tasks.
- Search & RecSys: Deep expertise in building and scaling Large-scale Search systems, Ranking (Learn to Rank), and Recommendation engines using ML/DL, productionized conversational bots
- Computer Vision & Multimodal: Deep understanding of Classical Computer Vision (OCR, segmentation, filtering) and modern Image Understanding models (CLIP, ViT). Hands-on experience with stable diffusion models is preferred.
- GenAI & Agents: Fine-tuning SLMs, training SLM/LLM from scratch, understanding of SOTA LLM models and capabilities, Multi-agent workflows (e.g., LangGraph, CrewAI, AutoGen), and AI-native "Co-pilot" architectures.
- Engineering Excellence: Strong system design skills for low-latency, high-throughput platforms.
- Experience with Vector DBs, Graph Knowledge Bases, and service layer architectures.
- Proficiency in coding skills with prior experience in Pyspark, Python, Tensorflow, Torch and version control systems.Strategic & Product Thinking
- Optimization: Ability to evaluate trade-offs between proprietary LLMs vs. fine-tuned open-source models regarding cost, latency, and accuracy.
- ROI Focus: A "Founder Mentality" that connects model performance to business KPIs and customer experience.
- Collaboration: Experience partnering with Product Managers to translate complex workflows into technical roadmaps.
- Expertise in evaluating build-vs-buy options and supporting recommendations with logical explanations.
Qualifications
- Experience: 6-9 years in AI/Data Science, with a proven track record of deploying models in production.
- NLP: Hands-on experience with transformer-based open-source SLMs and fine-tuningEncoder models (e.g., Nomic, RoBERTa, GTE etc.,) for specialized tasks.
- Search & RecSys:
- Deep expertise in building and scaling Large-scale Search systems, Ranking (Learn to Rank), and Recommendation engines using ML/DL, productionized conversational bots
- Computer Vision & Multimodal:
- Deep understanding of Classical Computer Vision (OCR, segmentation, filtering) and modern Image
- Understanding models (CLIP, ViT).
- Hands-on experience with stable diffusion models is preferred.
- GenAI & Agents: Fine-tuning SLMs, training SLM/LLM from scratch, understanding of SOTA LLM models and capabilities, Multi-agent workflows (e.g., LangGraph, CrewAI, AutoGen), and AI-native "Co-pilot" architectures.
- Engineering Excellence: Strong system design skills for low-latency, high-throughput platforms.
- Experience with Vector DBs, Graph Knowledge Bases, and service layer architectures.
- Proficiency in coding skills with prior experience in Pyspark, Python, Tensorflow, Torch and version control systems.
- Strategic & Product Thinking Optimization:
- Ability to evaluate trade-offs between proprietary LLMs vs. fine-tuned open-source models regarding cost, latency, and accuracy.
- ROI Focus: A "Founder Mentality" that connects model performance to business KPIs and customer experience.
- Collaboration: Experience partnering with Product Managers to translate complex workflows into technical roadmaps.
- Expertise in evaluating build-vs-buy options and supporting recommendations with logical explanations.
Responsibilities
- Architect Hybrid Systems:
- Build systems that leverage both Generative AI (Agents/LLMs) and traditional ML (Ranking/Search) to solve complex search and discovery problems.
- Scale High-Performance Models:
- Design and train large-scale Machine Learning and Deep Learning models for Swiggy's recommendation and ranking pipelines.
- Innovation & Benchmarking: Stay ahead of the curve by benchmarking emerging trends and contributing to the global AI community through research or blog publications.
- Cross-functional Leadership:
- Lead end-to-end projects from problem statements to production, ensuring scalable and cost-effective AI deployments.