About the role

Role Overview

Quality Assurance Engineer for a remote, full-time position in the United States. You will help develop, evaluate, and refine AI-driven systems where large language models operate as intelligent, multi-step agents delivering structured, actionable outcomes.

Responsibilities

Develop, test, and improve AI agent workflows in real-world scenarios involving multi-stage and iterative interactions.
Evaluate complex system behaviors with strong analytical rigor and attention to detail.
Ensure production-quality software design with clearly separated components (e.g., data ingestion, processing logic, and output/reporting).
Work on systems operating in live environments and manage persistent states across sessions.
Identify and help mitigate subtle vulnerabilities (e.g., data privacy risks, unauthorized access escalation, and prompt injection).

Requirements

Minimum 2 years of experience in backend development, AI automation, or systems integration.
Strong backend engineering background and ability to build production-quality software with separated components.
Proficiency in at least two programming languages such as Python, JavaScript, Go, or Java.
Hands-on experience working with SQL databases.
Experience building systems in live environments with multi-step or iterative interactions.

Preferred Skills

Experience designing multi-stage workflows where data collection feeds structured reasoning and decision-making.
Familiarity with integrating AI agents with external tools/platforms (e.g., APIs, databases, productivity tools like Supabase or Gmail).
Comfort with persistent system states and tracking progress across sessions.
Ability to spot and mitigate security risks including privacy and prompt injection issues.

About SoTalent

SoTalent is a talent and recruiting firm connecting candidates with companies building advanced technologies. This role is with a client focused on enhancing AI agents using human feedback and real-world system design, including training large language models to plan, coordinate, and execute complex workflows.