Program Manager, AI Agent & Support Automation

Bullhorn•Remote NOAM - Massachusetts, MA

About The Position

This is a builder role at the intersection of AI, customer experience, and operational excellence. You'll be the single accountable owner of Sophia, Bullhorn Support's customer-facing self-service AI agent (built on Salesforce Agentforce), driving measurable case deflection while keeping the customer experience excellent. You'll inherit strong foundational tooling but will need to diagnose gaps, tune the agent, establish guardrails and governance, and build the measure-and-improve cadences that make self-service durable at scale. This role owns and tunes the system; it doesn't require building it from scratch or writing production code.

Requirements

4+ years in support operations, customer experience, conversational/AI design, or a self-service/deflection role at a SaaS or technology company, including hands-on ownership of a self-service or AI support capability.
Hands-on experience configuring, deploying, and tuning conversational AI agents in production (Agentforce strongly preferred), including Agent Script.
Prompt engineering experience: designing, testing, and iterating prompts to improve accuracy and behavior.
Proven case deflection results using self-service AI tooling, with the ability to use data and conversation analysis to evolve self-service over time.
Salesforce Service Cloud (case management) and Experience Cloud (customer portal) experience.
Strong analytical skills, comfortable building dashboards/scorecards and translating metrics into recommendations.
Excellent written communication, including executive-level summaries.
Ability to operate as an individual contributor while influencing across functions, with product/project and change management experience.

Nice To Haves

Salesforce Administrator and/or Agentforce Specialist certification.
Low-code/scripting ability (Flow, APIs, JSON); builders and "vibe coders" welcome, production code not required.
Familiarity with KCS/knowledge management and how article quality affects AI deflection.
Experience with other AI-assisted support tools and BI/data-visualization tools (e.g., Snowflake, Tableau, Salesforce reports).
Background in staffing, HR tech, or enterprise SaaS, and experience supporting global teams across US and EMEA time zones.

Responsibilities

End-to-end ownership of Sophia: configuration, prompting, performance, and success, including subagents, topics, actions, routing, escalation-to-case, and guardrails.
Designing, testing, and iterating the prompts and instructions that drive the agent's behavior, tone, and accuracy, validated in sandbox with regression testing before every production release.
Case deflection and containment targets: defining them, driving them, and reporting on them, while protecting CSAT.
Production monitoring and QA: reviewing transcripts, tracking accuracy and drift, and catching regressions early.
Guardrails, human-in-the-loop policies, and responsible-AI practices, with a rollback plan for every change.
Connecting internal and external self-service, and representing Support's needs into Bullhorn's internal AI knowledge initiative.
Partnering across Support Operations, CX Data & Analytics, Enablement, and platform/vendor contacts, and providing executive-level updates on self-service performance.