About the Role
At Tabby, the LLM (Large Language Model) Agent handles most transactional contacts. What reaches a human is the hard part: judgment, research, empathy, and landing on a resolution the customer accepts as fair.
This role owns how well we do that. You define what a good customer conversation looks like and you build the capability across our managers to deliver it consistently. One question sits behind every conversation we look at: did we exhaust the opportunity to serve this customer.
The work happens through our Leaders. You build their ability to see the missed opportunities inside their own teams' conversations, and you hold them accountable for acting on them.
Customer Centricity Standard
- Define what a good customer conversation looks like and keep it to a short set of criteria.
- Evaluate on a single judgement per interaction: did it meet the standard, and if not, which of three to five reasons explains it.
- Keep the standard current as products, policies, and customer journeys change.
Manager Enablement
- Set the customer centricity standard and equip Team Leaders and Operation Managers to apply it with their teams.
- Run ticket-solving sessions with managers to surface the opportunities they walk past.
- Assess how Team Leaders coach their agents, using LLM analysis.
- Coaching of agents sits with the Team Leaders. This role makes sure that coaching is good.
CSAT
- Own the diagnosis of CSAT: what drives it, where it breaks, and who fixes it.
- Build the ability to recover CSAT even on outcomes the customer will not like.
- Split detractors into skill, will, or product and uncontrollable, so every point has a cause and an owner.
- Report blended CSAT and controllable CSAT as two separate numbers.
- Route product-driven dissatisfaction to product owners via the Experience Leads.
LLM Quality
- Own the LLM Quality Team, which runs checklists, per-conversation evaluation, and structured feedback into LLM agent “guidance” or prompt.
- Specialize evaluators by product stream and pair them with the Experience Leads.
- Maintain the LLM evaluation model separately from the human one, where LLM Agent checkers are more detailed, while human quality is outcome driven.
Analysis & Insight
- Work across everything we have on the customer: evaluation results, CSAT and DSAT verbatims, VoC, complaints, escalations, repeat contacts, and the conversations themselves.
- Dissect the data, and work with Analytics and with the analyst in the function to get what you need.
- Turn findings into refreshers, training triggers, and process changes with named owners and dates.
- Measure the function on whether a driver actually moved.
- Build a small team of high-skill owners with real accountability for outcomes.
- Experience coaching people to win customers over, in telesales, retail, hospitality, or service recovery.
- Not so much a checklist owner. Rather a sales coach. We have the product that we have to sell - the agents must be fluent in making the customers satisfied with the outcome regardless of excuses.
- Data fluent. You can take a large dataset, dissect it, and explain what it means without someone translating it for you. Comfortable working across large datasets and dashboards.
- Calm and measured.
- A track record of leading change that met resistance and made it stick.
- Able to hold a position with senior stakeholders on evidence, and change it when the evidence changes.
- 7+ years in CX, service, or sales coaching leadership, with real ownership of outcomes.
- Comfortable working in an LLM-heavy environment.
