AI Trainer

Okta Okta · Enterprise · Toronto, ON · Customer Acceleration-731

Okta is seeking an AI Trainer and Optimization Analyst to own and enhance their customer support AI agentic experiences. This role involves prompt engineering, performance analysis, and ensuring the AI agent provides accurate, on-brand, and ethical support, managing guardrails against hallucination. The goal is to optimize the agent's effectiveness and customer impact.

What you'd actually do

  1. Own Conversational Quality & Tone: Act as the primary owner for the AI agent's output, managing the content, tone, and ethical guardrails to ensure responses are accurate, on-brand, and avoid inappropriate or incorrect assumptions and "hallucination."
  2. Prompt Engineering & Training: Develop, test, and deploy advanced prompt engineering techniques, training materials, and conversational flows to enhance the AI agent's accuracy, intent recognition, and overall effectiveness.
  3. Performance Analysis & Optimization: Proactively monitor and analyze AI agent performance metrics, including deflection rates, resolution rates, and customer satisfaction (CSAT) scores, translating data into actionable training and optimization plans.
  4. Gap Identification & Remediation: Review and categorize high-volume AI conversations to identify new customer intents, common misclassifications, and critical content gaps in the knowledge base that require remediation.
  5. Strategic Collaboration: Work closely with the Content, Knowledge, and Customer Experience teams to understand common customer issues and translate those insights into strategic AI training plans, focusing on high-ticket volume areas.

Skills

Required

  • Proven experience in AI training, prompt engineering, Natural Language Processing (NLP), or conversational design for customer support or enterprise service systems.
  • Strong analytical skills with ability to interpret conversational data and performance metrics.
  • Familiarity with major AI conversational platforms (e.g., Dialogflow, IBM Watson, Rasa, or similar enterprise-level tools).
  • Excellent written communication skills.
  • Ability to understand the broader online customer support ecosystem.
  • Strategic mindset and problem-solving approach.

Nice to have

  • Experience with AI agent optimization
  • Experience with customer support journeys

What the JD emphasized

  • primary expert
  • ownership
  • hallucination
  • prompt engineering
  • conversational data
  • performance metrics
  • customer support
  • enterprise service systems
  • conversational AI platforms

Other signals

  • AI agent performance analysis
  • prompt engineering
  • conversational quality and tone management
  • customer support optimization