Note: The job is a remote job and is open to candidates in USA. Zearch is partnering with a Series C AI/SaaS business building an Agentic Customer Experience platform. The Principal Product Manager will own the AI support system and operating model, improve automation and customer experience metrics, develop a senior team, and use support insights to shape the product roadmap.
Responsibilities
- The AI support system, as a product — knowledge base, AI-driven triage and responses, containment targets per ticket category, quality evals, auditability, and clean escalation to humans. You build and train the system yourself; like any product, it carries its own roadmap
- Support operating model — the structure, workflows, swim lanes, escalation paths, and governance spanning each customer segment (self-serve/new-to-platform, integrations, mid-market, enterprise), each with a dedicated playbook
- Deflection through automation and root-cause fixes — contact rate is treated as a product defect metric. You'll track tickets per active customer by segment, surface the leading root causes each quarter, and eliminate them at the source, whether that's documentation, in-product changes, self-serve tools, or better error messaging
- Voice of the customer as a signal engine — you'll consolidate fragmented feedback channels into one structured, revenue-weighted pipeline, and co-own a quarterly "top ticket-driver elimination" list with Product, tracking what actually ships and how much volume it removes
- Team leadership — building a small, senior organization complete with a career ladder, individual development plans, and a real review and merit process. A year in, you should be able to point to people who grew because of you
- Service levels and quality — accountable for TTFR, TTR, and CSAT, backed by weekly scorecards, trend analysis, and corrective actions
- Capacity planning — forecasting ticket volume and complexity by segment, then staffing and scheduling across time zones (including EU coverage) without treating headcount as the default lever
- Premium support tiers — designing and running paid offerings that scale without turning into a dedicated-services model
- Governance of managed hacks — owning a visible system for tracking the scripts, automation flows, and workarounds the support team creates, with durable ones pitched back to Product as features you champion
- Design and operate the support model — swim lanes, handoffs, escalation playbooks, and a playbook library organized around the customer journey — to cut variance and raise first-contact resolution
- Continuously improve AI-driven resolution, covering accuracy, containment, safety, auditability, and human escalation for edge cases, using evals you design rather than dashboards inherited from a vendor
- Take ownership of contact rate and deflection results, running your own analysis — ticket categorization, cohort work, dashboards — since this role builds its own instruments rather than relying on someone else's
- Set and manage SLAs and weekly scorecards, following through with corrective action plans
- Develop the team through coaching, development plans, mid-year and annual reviews, merit decisions, and a career ladder that turns analysts into leads
- Operate the revenue-weighted voice-of-customer pipeline, feeding quantified ticket-driver data into product discovery, release planning, and prioritization discussions, and reviewing upcoming releases for supportability
- Model capacity and staffing needs across time zones based on forecasted volume by segment and stage
- Launch premium support tiers along with the packaging that supports them
- Build out the managed-hacks register, covering ownership, documentation, risk review, lifecycle management, and the route to productization
Skills
- A hands-on builder who has personally built pieces of a support system before — AI workflows, prompts and evals, dashboards, automations — and uses AI daily as part of the job. This is a role for someone who builds the system, not just runs the queue
- A proven people developer , with concrete examples of ICs you've grown into leads or leaders and how you did it. This is a hard filter: exceptional individual contributors who prefer doing everything themselves won't succeed here
- Strong analytically , bringing numbers to the table without being asked — containment by category, contact rate by cohort, cost per resolution — and capable of building the dashboards yourself
- Experienced in change management under real constraints , having driven adoption of tools like Intercom-class platforms and documented tribal knowledge inside an organization resistant to both
- Comfortable with startup pace and ambiguity , figuring out what needs doing without waiting for a plan to be handed to you
- A track record of treating support as a product feedback engine , having generated quantified signal, shaped roadmaps, and measured the ticket reduction that resulted from shipped changes
- Experience leading support in technical B2B settings — integrations, APIs, production incidents — where the conversation is engineer to engineer
- Background in a technical, integration-heavy B2B SaaS space (APIs, data pipelines, workflow automation, or similar) — deep expertise isn't required, but the ability to get dangerous in the domain within a quarter is
- A track record of running or building support at a company scaling revenue or volume 2x+ annually
- Hands-on experience deploying a Fin/Sierra/Decagon-class AI agent, with real containment metrics and eval rigor to show for it
- Experience designing premium or paid support tiers without them turning into a services dependency
Company Overview