Senior AI Program Director for Agent Platform Delivery

Programmers.io; Sunnyvale (CA); On-site; USD 190,000 - 230,000; Full time

Job summary

Programmers.io seeks a Senior Technical Program Manager to own the operating model of AEGIS, an AI agent platform delivering governance, data products, and multi-agent coordination for global sales analytics. The role requires steering release governance, evaluation gates, and cross-team collaboration in a high-velocity environment.

You will inherit an active program mid-flight, stabilizing and improving it while driving quality signals, prompt changes, and model-versioning considerations for

Qualifications

• 10+ years leading complex software, data, or platform programs.
• Shipping LLM-based or ML products to production with agent tooling.
• Experience designing triage, quality ops, with measurable improvements.
• Reasoning about APIs, data contracts, and CI/CD pipelines.
• Executive presence with senior leadership reviews.
• Ability to drive in-flight programs within weeks.

Responsibilities

• Design end-to-end delivery rhythm across engineering pods.
• Maintain a dependency-aware roadmap with named owners.
• Own release readiness and go/no-go decisions with cross-functional leads.
• Manage intake-to-resolution pipeline for defects and data gaps.
• Instrument triage metrics and redesign handoffs to reduce stalls.
• Lead root-cause analysis for quality incidents and remediation.

Skills

Technical program mgmt

ML product experience

AI agent platforms

Cross-team leadership

Executive communication

Release governance

Data governance basics

Tools

Langfuse

Arize Phoenix

LangSmith

Snowflake

Job description

Programmers.io seeks a Senior Technical Program Manager to own the operating model of AEGIS, an AI agent platform delivering governance, data products, and multi-agent coordination for global sales analytics. The role requires steering release governance, evaluation gates, and cross-team collaboration in a high-velocity environment.

You will inherit an active program mid-flight, stabilizing and improving it while driving quality signals, prompt changes, and model-versioning considerations for