About the role
Job Description A bit about this role: As a Senior Manager, AI Performance & Operations, you will own the production performance and continuous improvement of Devoted’s core agentic workflows for service and experience. so that every general-availability workflow has clear performance thresholds, reliable monitoring, explicit fallback paths, and a measurable improvement plan. This person is the accountable operator for Guide’s highest-volume and highest-risk AI workflows once they are live. They maintain the portfolio view of where each workflow stands, detect degradation early, diagnose the root cause, drive corrective action, and verify that implemented changes improve outcomes. They also serve as the in-house expert for early-stage projects by helping teams define instrumentation, evaluation design, launch criteria, and safe hand off into general availability.
Responsibilities and Impact will include: Post-Launch Accountability: Own the health of major voice and text AI workflows, ensuring business, quality, safety, and operational outcomes are met. Portfolio Oversight: Maintain a current view of all major workflows, including status, maturity, failure modes, and next actions. Performance & Dashboards: Monitor weekly indicators (containment, escalation, complaints) and oversee production dashboards to track behavior independently. Diagnosis & Incident Response: Lead response for AI regressions; distinguish between model, prompt, or orchestration issues and drive post-mortems to closure. Continuous Improvement & Optimization: Run structured cycles (hypothesis, testing, rollout/rollback) for "mini-launches" like model swaps, prompt changes, and tool updates. Recommend evidence-based accelerations, such as model migrations or deeper tool-orchestration changes. Scalable Evaluation: Design eval systems using golden datasets, LLM-as-judge workflows, and human review queues for production and pre-production. Expert Review: Guide new workflows during incubation to ensure they launch with proper instrumentation, eval coverage, and safety models. Executive Communication: Translate technical changes into plain language for leadership, explaining what broke, why it matters, and who is fixing it. Required skills and experience: Applied AI Expertise: Strong production experience with LLMs, RAG, voice AI, agentic orchestration, or equivalent ML systems. Evaluation Systems: Demonstrated experience designing or operating rubrics, scoring frameworks, regression tests, and production-sampling workflows. Diagnostic Acumen: A clear mental model for distinguishing failure modes, such as model reasoning, tool failure, retrieval issues, and workflow logic. Optimization Mindset: Comfort with active experimentation—including prompt/context tuning, retrieval redesign, and safe model migrations—beyond simple issue triage. Strategic Communication: Excellent cross-functional judgment and the ability to communicate clearly in ambiguous, high-accountability environments. Outcome Ownership: Deep personal accountability with a drive to own results rather than just producing recommendations. Preferred Industry Background: Experience in healthcare, regulated workflows, contact center operations, or high-stakes service quality/reliability functions. Salary: $110,000 - $172,000 / year The pay range listed for this position is the range the organization reasonably and in good faith expects to pay for this position at the time of the posting. Once the interview process begins, your talent partner will provide additional information on the
