Senior Director, Technical Product Management — AI Quality & Evaluation

What You’ll Be Doing & Why It Matters

• Help build the next generation of Agentforce.
• We’re an entrepreneurial incubation team within Salesforce, combining the pace and ownership of a startup with access to enterprise customers and Salesforce’s distribution.
• We’re looking for a technical product leader with deep AI expertise who wants to turn frontier capabilities into products customers can rely on.

Our team works closely with AI researchers, applied scientists, and engineers, uses coding agents extensively, and builds and experiments outside the core platform to iterate quickly. You’ll shape product strategy for AI quality and evaluation, establish how we measure progress, and personally lead a customer quality workstream and a focused feature from early learning through delivery.

• This is a hands-on individual contributor leadership role, with an IC focus expected through 2027.
• You’ll lead through product ownership, technical judgment, and cross-functional influence while coaching colleagues, including junior designers and technical program managers.
• People-management responsibilities may develop as the product and team grow.
• You should enjoy inspecting traces and data, challenging evaluation results, prototyping with coding agents, and working directly with customers.
• Engineering partners own technical architecture and production implementation.

Core Responsibilities & Scope

• Set a coherent product strategy and roadmap for AI quality and evaluation, connecting customer needs, technical opportunities, and measurable business outcomes.
• Establish shared quality definitions, evaluation requirements, coverage priorities, and release criteria with research, engineering, and product partners.
• Make explicit investment tradeoffs across customer quality improvements, evaluation capabilities, and product features; align teams on priorities and decision ownership.
• Personally lead customer investigations and experiments. For example, investigate why an agent gives an incorrect answer or fails a tool-driven workflow, define the expected behavior, and partner with researchers and engineers to test a change and validate its customer impact.
• Build an operating approach that connects customer feedback, evaluation evidence, experiments, releases, and learning across teams.
• Partner with researchers, applied scientists, and engineers to design credible evaluations. Apply AI/ML fundamentals such as training/test separation, overfitting, representative coverage, and independent validation to assess non-deterministic model behavior and grader reliability.
• Lead delivery of capabilities that help customers understand and improve AI performance. For example, prototype an experience that turns a failed agent interaction into an evaluation case, makes the quality gap clear, and helps a customer verify whether a proposed change works.
• Coach product partners, junior designers, and technical program managers in problem framing, experimentation, and delivery; raise the quality of decisions through mentorship and practical example.
• Bring verified competitive insight into strategy and communicate differentiation, limitations, and investment choices clearly to customers and senior stakeholders.