Responsibilities
THE TEAM:
We are a small, AI-forward data science team with two jobs. We pull commercial data out of Salesforce through MCP into Snowflake and turn it into AI dashboards built for the question each audience is actually asking, and we own the shared library where AMD teams publish the AI skills and agents they build. The dashboards are where we prove AI-assisted analysis holds up against real, high-stakes data; the library is how that work stops being ours alone.
THE ROLE:
AMD is seeking a Principal Program Manager to own the operating model across both halves of that charter—the standards, decision rights, cadence, and measurement that turn scattered good work into a system.
You will own the dashboard portfolio as a product rather than a request queue: who each surface is for, what decision it drives, what gets built next, and what gets retired. Co-equal with that, you will own the trust layer underneath—definitions, named owners, pipeline reliability, and the honest answer when a number is wrong. On the library side, you will define how reusable AI assets are contributed, reviewed, and published, charter the councils that decide, and own the promote-to-enterprise path to peer AI Libraries across AMD.
This is a senior individual contributor role with a builder bench alongside you, not underneath you. Hands-on skill—SQL, Python, Snowflake, dbt, dashboards, or agent engineering—is welcome and will make you more effective. It is not a requirement.
THE PERSON:
You have moved an organization from AI enthusiasm to AI operating behavior, and you know the difference between deployment and adoption. You are fluent enough in how skills, agents, and prompts are built to set a quality bar engineers respect, and you would rather partner with builders than out-argue them. You bring editorial judgment to a domain where it is trivial to produce more dashboards and more assets than anyone will use.
KEY RESPONSIBILITIES:
- AI Dashboards & Decision Products
- Own the dashboard portfolio as a product: audience, decision it drives, roadmap, and what gets retired.
- Differentiate the IC, manager, and executive views—each answering a different question, not one dataset shown three ways.
- Define where AI belongs in a dashboard—summarization, anomaly surfacing, natural-language query—and where it does not.
- Write the requirements builders implement: metrics, grain, filters, refresh, and acceptance criteria.
- Data Foundation & Trust
- Own metric definitions and the named owner behind each one, so two dashboards cannot disagree without someone accountable.
- Set the quality bar for the Salesforce-to-Snowflake path, and partner with builders and IT on pipeline monitoring and validation.
- Run the triage path when a number is disputed: find the truth, publish the correction, and fix the cause upstream.
- AI Library & Review Councils
- Define and operate the asset lifecycle: submission, review, quality bar, versioning, publication, and retirement.
- Charter and run the councils—membership, intake, pre-read, decision rights, escalation—and build the criteria-based go/no-go framework they decide against.
- Own the promote-to-enterprise path, including the security, privacy, and Responsible AI review gate.
- Adoption & Measurement
- Design the adoption architecture: phased onboarding, rollout sequencing, champion engagement, and support model.
- Measure adoption on consistent usage, active cohorts, and value proxies—not license counts—and report it on a shared scorecard, including what is not working.
- PREFERRED EXPERIENCE: Program, transformation, or operations leadership at enterprise scale, turning executive direction into governance, cadence, and measurable outcomes.
- Ownership of an AI or digital adoption program: rollout sequencing, champion networks, readiness criteria, adoption measurement.
- Experience designing operating models—decision rights, intake and approval workflows, escalation, executive review cadence.
- Ownership of a reporting portfolio serving ICs, managers, and executives, including the discipline to retire what stopped earning its place.
- Working knowledge of Salesforce commercial data and a modern cloud warehouse, and of what breaks between source system and dashboard.
- Fluency with generative AI—LLMs, agents, prompting, evaluation—enough to set a credible quality bar with technical contributors.
- Strong executive communication, influence without authority, and semiconductor or comparable industry experience are all a plus.
- Recommended, not required.
- Hands-on depth in SQL, Python, Snowflake or dbt, dashboard development, MCP and API integration, or agent engineering is weighted favorably but is not a screening criterion.
- ACADEMIC CREDENTIALS: Bachelor's or Master's degree in Engineering, Information Systems, Business, or a related field, or equivalent experience.
- LOCATION: Austin, TX
- This role is not eligible for visa sponsorship.
Qualifications
