SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salesforce is hiring a Director of Product Management to own agentic analytics for Agentforce, the company's platform for building, evaluating, and managing AI-powered agents. This is a high-leverage role focused on helping enterprises measure, monitor, and improve their AI agents in production.
In this position, you will own the complete product lifecycle for agentic analytics—from current pilot phase through general availability and beyond. You'll set and defend the roadmap, outcomes, and product strategy. Key responsibilities include owning the user experience in partnership with design, managing go-to-market strategy with enablement, docs, marketing, and sales, and running a tight design partner and pilot program with early customers.
You'll be expected to prototype with coding agents before writing specs, dogfood the product constantly against real customer data, and stay close to the code and runtime. You'll run the design partner program directly, lead pilot calls, extract requirements, and turn customer pain into shipped product. Success metrics include pilot activation, failure-cluster coverage, recommendation acceptance, and agent quality lift.
This role requires deep product management experience shipping products for technical users (developers, admins, technical SMEs), with a clear pattern of scope and impact. You must be a daily user of coding agents with hands-on experience building and shipping agents yourself. You should have applied depth in evaluation frameworks, annotation flows, failure clustering, and the ability to reason about measuring agent quality improvements.
You'll work across engineering, design, and adjacent product teams, turning fuzzy problems into sharp bets with measurable outcomes. Strong written and verbal communication is essential—you'll rally teams around direction and hold your own with customers and executives. An engineering background or equivalent hands-on depth is preferred, including comfort reading code, reviewing APIs, and prototyping. Direct experience with the current agent-building and evaluation landscape (Cursor, Claude Code, Replit, LangSmith, Braintrust, Langfuse, agent frameworks) and strong opinions on their strengths and weaknesses is highly valued.
