> Markdown version of [/jobs/ext/4618377-ai-engineer-4](https://www.wearedevelopers.com/jobs/ext/4618377-ai-engineer-4). 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).

---

# AI Engineer 4

- **Company:** Adobe Inc.
- **Location:** San Jose, CA, United States
- **Experience:** Experienced
- **Contract:** Permanent contract
- **Skills:** Artificial Intelligence, Amazon Web Services, Microsoft Azure, Continuous Integration, Python (Programming Language), Machine Learning, Neo4j, Data Streaming, Large Language Models, Apache Spark, Llamaindex, Backend, Pgvector, Kubernetes, Apache Flink, Apache Kafka, Claude, Restful APIs, Data Pipelines, Docker, Databricks
- **Published:** November 7, 2026
- **Apply:** https://www.juju.com/job/28_25303_019c4427d15c38d0cfa4c823fac46517

## About the Role

* Around 3+ years building AI/ML or backend systems, including some experience running LLM-powered services in production.
 * Strong fundamentals in Python and REST APIs.
 * Familiarity with modern AI tooling - frameworks like LangChain or LlamaIndex, and vector databases.
 * A habit of measuring quality and iterating, rather than shipping and hoping.
 * Comfort working with cloud platforms (AWS, GCP, or Azure) and containers (Docker, Kubernetes)., * Experience with event streaming (Kafka, Flink, or Kinesis) and data pipelines (Spark or Databricks).
 * Exposure to production operations - CI/CD, monitoring, alerting, and incident response.
 * Interest in AI governance, safety, or evaluation.

## Description

* Build and productionize reusable agentic components - skills, orchestration workflows, and tool-calling integrations - that plug into OneAI's intelligence layer (Neo4j + pgvector + Databricks + Claude).
 * Take AI models and agents from prototype to production, and keep them healthy once they're live.
 * Improve how agents reason: prompt design, memory and context management, retrieval quality, and multi-step tool use.
 * Tune for latency, reliability, and cost so the platform holds up as more teams rely on it.
 * Partner with product managers and data engineers to shape what we build, and share what you learn with the team.
 * Help raise the bar on evaluation - measure solve rates and accuracy, and use what you find to make the agents better.