Perplexity is seeking an experienced Software Engineer to help build the next-gen AI Foundation & Platform. In this role, you'll help build Perplexity's end-to-end AI data, evaluation and personalization infrastructure and flywheel which powers almost all agent products.
What you'll do
- Collaborate closely with AI Product, Applied ML, Post-Training, and Data Science teams to design, build, and maintain scalable data pipelines and data lakes
- Develop high-performance infrastructure that powers personalization features including memory, discover, and agentic products
- Create a scalable, multi-modal evaluation platform for all Perplexity AI products, including personalization, pro search, labs, deep research, and Comet
- Design tools and abstractions on foundational infrastructure to enhance personalization, analytics, recommendations, AI products, and post-training capabilities
- Holistically improve engineering foundation to support rapid growth of Perplexity products and international user base.
What you need
- Strong programming and data engineering skills, with proficiency in open source & distributed framework(AWS, Spark, Flink, Iceberg, DynamoDB)
- Familiarity with cloud-based data services (e.g., AWS, RDS, DynamoDB), containerized infrastructure (e.g., EKS, Docker), and data streaming (Flink, Spark streaming, CDC)
- Strong quantitative and engineering skills with experience in estimating performance at high scale
- Experience supporting various ML/AI engineering teams to build scalable frameworks to accelerate R&D for frontier models and AI products
- Experience iterating on improving LLM responses and set up proper evaluation framework or Judges to analysis performance holistically.
- Self-motivated with a strong sense of ownership of systems and designs
- 5+ years of industry experience in distributed systems or AI infrastructure
