Drimva

Amplitude

Principal Product Manager, AI Agents & MCP

$237K–$398K

San Francisco · Principal · Full-time · Posted Aug 25, 2026

Experience
7+ years of product management experience, including 0-to-1 product delivery; 3+ years building products on AI/ML technologies.

About the job

Lead product strategy and delivery for Amplitude’s AI agents and MCP capabilities, helping customers gain insights and act through the product interface and external tools.

What you'll do

  • Set the 12–24-month direction and investment priorities for agents and MCP.
  • Lead product vision and collaborate with engineering and design to deliver agentic products and improve their quality using evaluations and feedback.
  • Engage customers and enable field teams to support adoption and use of AI Agent products.

What they ask for

  • 7+ years of product management experience, including taking 0-to-1 products from concept through delivery.
  • 3+ years building products on AI/ML technologies, with recent hands-on experience delivering agentic products.
  • Fluency in current LLM trends, models, and tools, including context management, latency, evaluation methods, and common failure modes.
  • Experience measuring whether a probabilistic product improves over time.

Pay

$237,000 - $356,000 USD total target cash (inclusive of bonus or commission) (Colorado); $265,000 - $398,000 USD total target cash (inclusive of bonus or commission) (New York City); $265,000 - $398,000 USD total target cash (inclusive of bonus or commission) (San Francisco Bay Area of California); $237,000 - $356,000 USD total target cash (inclusive of bonus or commission) (California outside of the San Francisco Bay Area); $237,000 - $356,000 USD total target cash (inclusive of bonus or commission) (Washington state)

Tools

MCP, LLM

AI focus

Set strategy and lead development of AI agents and MCP capabilities; establish evaluations, quality standards, and feedback mechanisms to measure and improve customer value.

Model-written summary · the original posting is authoritative

First seen Oct 27, 2026 · Last verified Nov 6, 2026 (UTC)