AI Engineer

The IBM Customer Zero initiative builds internal AI tooling that improves how the IBM workforce operates. This role focuses on the sales motion: building applications that help IBM sales teams identify, interact with, and assist customers and prospective customers. The work draws on IBM's distinctive position in the AI market, including the ability to build models in the Granite series, host models across IBM infrastructure, and apply IBM software in the construction of AI-centric and AI-enhanced systems. This is a hands-on engineering role located in Austin, with Atlanta as a secondary location.

Your role and responsibilities

As an AI Engineer on the IBM Customer Zero initiative, you will build internal AI-powered tools that help IBM sales teams identify, engage, and support customers and prospective customers across the IBM software portfolio. You will work directly with sales stakeholders to understand their workflows, then design and ship working applications that improve those workflows in measurable ways.

Your primary responsibilities will include:

• Design and build AI-powered applications. Architect and deliver production-grade tools that apply current agentic frameworks, coding agents, and large language model APIs to real sales workflows. Ship working software, not prototypes.
• Work directly with sales stakeholders. Observe sales workflows, identify high-impact problems, translate those observations into requirements, and validate solutions against real use. Communication skill matters as much as engineering skill in this role.
• Apply the full model spectrum. Build solutions that draw on the appropriate model for the task, including IBM Granite, open-weight and open-source models, and current frontier models from major providers. Understand the tradeoffs between them.
• Operate effectively in coding-agent-driven engineering. Contribute to a codebase where agents accelerate development substantially, including the disciplines required to manage merge volume, code collisions, and consistency drift that come with that pace.
• Establish evaluation and measurement discipline. Define how each tool will be measured before it is built. Build evaluation harnesses, monitor production behavior, and identify regressions before users do.
• Extend over time toward buyer-facing tooling. As the team matures, contribute to tools that help software buyers navigate solution design and the buying process. This is a roadmap direction, not a day-one responsibility.

Required technical and professional expertise

• Fluency in Python, with working knowledge of at least one additional modern language such as JavaScript or TypeScript. Several years of professional engineering experience.
• Fluency in React, with working knowledge of another modern framework such as Vue or Angular acceptable
• Hands-on experience building applications that use large language model APIs across multiple providers, including frontier models and open-weight or open-source models.
• Working experience with current agentic frameworks and coding agents, including the ability to assess which framework fits a given problem.
• Working knowledge of relevant AI and integration protocols, including standard API design, Model Context Protocol, and Agent-to-Agent communication. Candidates are not expected to have shipped production A2A systems, but should understand what A2A is and the problems it addresses.
• Working knowledge of vector search tooling and the database interactions that support retrieval at production scale.
• Working knowledge of CI/CD practices and the disciplines that keep modern application codebases deployable and consistent.
• Exposure to coding-agent-driven engineering at pace, including the practical reality of managing merge volume and code collisions when agents contribute substantial portions of a codebase.
• Demonstrated ability to define how an AI application will be evaluated, including the construction of evaluation datasets and the measurement of model and system behavior in production.
• Strong written and verbal communication, including the ability to translate between technical implementation and non-technical stakeholders.

Preferred technical and professional experience

• Familiarity with the IBM AI portfolio, including WatsonX and the Granite model family, or clearly demonstrated interest in developing that familiarity.
• Experience designing or contributing to internal tools used by non-technical users in production settings.
• Working knowledge of production AI engineering practices, including observability for model-based systems, cost monitoring, and graceful handling of model API failures.
• Familiarity with sales workflows, CRM data models, or pipeline mechanics, sufficient to collaborate effectively with sales stakeholders.
• Experience with containerization, infrastructure-as-code, and deployment practices common to modern application engineering.