Applied AI Engineer

Gradial•Seattle, WA

18d•$130,000 - $220,000

About The Position

Gradial is seeking an Applied AI Engineer to join their team and build the next generation of agentic systems. This role involves a blend of deep software engineering skills and an experimental approach to applying AI to solve user problems. The engineer will collaborate with product, research, and engineering teams to design agents capable of reasoning, interacting with environments, and evolving through feedback. The ideal candidate will possess curiosity, enjoy tackling ambiguous problems, and have practical experience deploying LLMs in production environments.

Requirements

4+ years of software engineering experience, with at least a year working directly on AI/ML-powered applications.
Strong fluency in AI tooling (e.g. OpenAI APIs, Anthropic APIs).
Practical understanding of LLMs, prompt engineering and agentic architectures.
Experience building production systems that integrate external environments - code execution, browser automation or API orchestration.
High degree of ownership and comfort in experimental, fast-moving environments.

Nice To Haves

Familiarity with benchmarks like SWE-bench or evaluation strategies for tool-using agents.
Exposure to customer-facing roles or user research for AI product development.
Background in reinforcement learning or data-centric AI approaches.

Responsibilities

Build autonomous and semi-autonomous agent workflows that interact with browsers, codebases, and APIs to complete complex content operations.
Evaluate model behavior in live environments and pioneer approaches for applying reinforcement learning to end-to-end agent systems.
Design modular frameworks for tools, memory, and evaluators to enable continuous learning and improvement of agents.
Develop data pipelines to collect, transform, and annotate real-world usage data for SFT and RL training.
Design and build proprietary agent libraries, prompting strategies and benchmarking frameworks.
Partner with customers and internal stakeholders to understand edge cases and align AI behaviors with user expectations.