> Markdown version of [/jobs/ext/4322590-applied-scientist-aws-agentic-ai](https://www.wearedevelopers.com/jobs/ext/4322590-applied-scientist-aws-agentic-ai). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Applied Scientist, AWS Agentic AI

- **Company:** Amazon.com, Inc.
- **Location:** Santa Clara, CA, United States
- **Experience:** Experienced
- **Salary:** $171,600.0 - $222,200.0
- **Contract:** Permanent contract
- **Skills:** Java (Programming Language), Artificial Intelligence, Amazon Web Services, Business Software, C++ (Programming Language), Static Program Analysis, Data Mining, Distributed Systems, Python (Programming Language), Machine Learning, Parsing, Software Engineering, High Performance Computing, Generative AI, Agentic-AI, Build Tools
- **Published:** October 28, 2026
- **Apply:** https://www.indeed.com/viewjob?jk=de4bb0d749d313f1

## About the Role

* 3+ years of building models for business application experience
 * PhD, or Master's degree and 4+ years of CS, CE, ML or related field experience
 * Experience in patents or publications at top-tier peer-reviewed conferences or journals
 * Experience programming in Java, C++, Python or related language
 * Experience in any of the following areas: algorithms and data structures, parsing, numerical optimization, data mining, parallel and distributed computing, high-performance computing

PREFERRED QUALIFICATIONS

 * Experience using Unix/Linux
 * Experience in professional software development

## Description

Great code assistants have to understand code the way an experienced engineer does, not as a wall of text but as a living structure of symbols, calls, types, and dependencies that span an entire repository. Most AI tools today flatten all of that into tokens and lose the structure that makes real software navigable. Amazon Web Services (AWS) is assembling an elite team of scientists and engineers to fix that, and we are hiring an Applied Scientist to build the intelligence layer that lets our AI development tools reason about code through its true structure. You will help create generative AI technologies, including Kiro IDE and CLI, that are transforming how software gets built.

You will design how we turn the many signals around code, from static analysis and symbol graphs to build systems and language servers, into a single unified representation our models can reason over. This is a build-oriented research role where the questions are still open: how to fuse heterogeneous code signals, how to represent a whole repository so a model can find and use the right context, and how to ground an assistant's reasoning in the structure of real software rather than text alone. Your work will reach millions of developers, and you will own this direction end to end, from framing the approach to shipping it in a product engineers rely on every day.

Key job responsibilities

 * Design a unified representation of code that fuses static analysis, symbol and call graphs, type information, and build system context

 * Build the methods and models that consume that representation to power code navigation, retrieval, and reasoning at repository scale

 * Invent and implement state-of-the-art machine learning solutions that operate at unprecedented Amazon scale

 * Deploy products that directly impact the daily workflows of millions of developers worldwide

 * Publish and present your work at premier ML, NLP, and software engineering conferences (NeurIPS, ICML, ICLR, ACL, EMNLP, ICSE, FSE)

 * Accelerate innovation by working directly with customers to rapidly transition research breakthroughs into production systems