Software Development Engineer, Recsai

Amazon Amazon · Big Tech · IL, Tel Aviv · Software Development

Software Development Engineer role focused on building and operating features for Amazon's next-generation recommendation system (RecsAI), which is built on a custom-trained LLM for shopping. The role involves designing and delivering end-to-end solutions at Amazon scale, integrating LLMs into real-time recommendation experiences, and collaborating with product and science teams.

What you'd actually do

  1. Design, build, test, and operate features for a personalized recommendation system used by multiple teams and operating at Amazon scale
  2. Deliver end-to-end solutions with focus on maintainability, scalability, performance, and reliability
  3. Collaborate with Product and Science to define experiences, run experiments, and iterate based on data
  4. Define and implement measurement strategies including analytics events and experiment configurations to track engagement and retention
  5. Navigate ambiguity and make sound technical decisions in a problem space where established patterns don't always apply

Skills

Required

  • Bachelor's degree in Computer Science, Engineering, Mathematics, or a related field
  • 5+ years of non-internship professional software development experience
  • Experience programming with at least one modern language such as Java, C++, or C# including object-oriented design
  • Experience with full software development life cycle, including coding standards, code reviews, source control management, build processes, testing, and operations
  • Experience contributing to the architecture and design (architecture, design patterns, reliability and scaling) of new and current systems

Nice to have

  • Master's degree or equivalent
  • Experience including, building and maintaining data flows and pipelines
  • Experience with A/B testing
  • Familiarity with AI/ML integration and generative AI applications
  • Experience with end-to-end SDLC ownership, including operations and on-call, monitoring/metrics, and incident response/RCA

What the JD emphasized

  • design and build features that shape how hundreds of millions of customers discover products
  • genuinely novel
  • defining how LLMs integrate into real-time recommendation experiences, not applying established playbooks
  • ship iteratively
  • real ownership over what they build
  • bias for action
  • ownership of problems end to end
  • operational excellence
  • making sure they hold up at scale
  • raises the bar for the team
  • mentoring junior developers
  • advocating for engineering best practices
  • thinking beyond the immediate sprint

Other signals

  • LLM trained specifically for shopping
  • intersection of personalization and generative AI
  • defining how LLMs integrate into real-time recommendation experiences
  • ship iteratively, and see your work in the hands of customers quickly