17 days ago
Bengaluru, IndiaStaff+
Responsibilities
- Design, build, and scale infrastructure and pipelines for online and batch machine-learning model serving across multiple modalities and workloads.
- Build production-grade services and APIs for internal and external products.
- Automate model training, evaluation, deployment, rollback, monitoring, and retraining workflows.
- Optimize inference latency, throughput, reliability, and cost through batching, caching, parallelism, quantization, and architecture-aware improvements.
- Improve engineering rigor through testing, observability, reproducibility, CI/CD, and incident response.
- Collaborate with product, platform, and software teams to turn ambiguous business problems into production ML systems.
- Mentor engineers and make pragmatic technical and architectural decisions while remaining a hands-on individual contributor.
Requirements
- 8+ years of experience in software engineering, machine learning engineering, or ML infrastructure.
- Strong experience building and operating production ML systems as a self-directed owner.
- Deep expertise in Python and solid backend engineering fundamentals, including APIs, distributed systems, testing, debugging, and operational ownership.
- Proven experience building production data or ML systems on AWS.
- Understanding of model-serving tradeoffs involving latency, throughput, autoscaling, concurrency, accelerator usage, memory pressure, and cost.
- Experience automating the ML lifecycle from experimentation and training through deployment and monitoring.
- Hands-on experience with PyTorch or equivalent modern ML frameworks.
- Ability to drive technical work end to end and make pragmatic architectural decisions.
- Strong communication skills and willingness to mentor engineers while remaining deeply hands-on.
- Preferred: experience serving and optimizing LLMs in production; computer vision, image understanding, vision transformers, or multimodal retrieval; Kubernetes, containerization, or large-scale distributed serving; inference optimization; vLLM, Triton, TensorRT, or similar inference stacks; and ML evaluation, observability, and monitoring workflows.
Tech Stack
Categories
About Nielsen
Nielsen shapes the world’s media and content as a global leader in audience insights, data and analytics. Through our understanding of people and their behaviors across all channels and platforms, we empower our clients with independent and actionable intelligence so they can connect and engage with their audiences—now and into the future. Nielsen operates around the world in more than 55 countries. Learn more at http://nlsn.co/6006JMfty and connect with us on social media (LinkedIn, Twitter, Facebook and Instagram).