AI Operations Specialist (Customer Support)

Sinch; Taguig; Hybrid; PHP 900.000 - 1.300.000; Vollzeit

Zusammenfassung

Sinch in Manila region is seeking an AI Operations Specialist focused on Customer Support. You will map intuitive self-service paths, define agent personas, and architect conversation logic across messaging channels.

Collaborate with Support Engineering and Product to automate workflows and ground AI answers in Sinch documentation. You will refine prompts, run A/B tests, and monitor CSAT, FCR, and containment metrics to improve customer outcomes.

Qualifikationen

• Ability to design multi-turn dialogue trees for enterprise/SaaS
• Strong prompt engineering and context management
• Experience using data to benchmark AI performance
• Understanding of support metrics (CSAT, deflection, AHT)

Aufgaben

• Map self-service conversation paths across channels from first touch to resolution.
• Define agent personas, tone, and empathy guidelines for different tiers.
• Architect conversational logic for ambiguous intents and edge cases.
• Develop prompts, few-shot examples, and dynamic context rules.
• Integrate AI agents with Zendesk and CRM systems.
• Collaborate with product and support teams to automate workflows.
• Review logs and optimize prompts based on CSAT and FCR data.

Kenntnisse

Dialog flows

Prompt engineering

LLM behavior

CSAT metrics

Conversation analytics

Tools

Zendesk

Jira Service Desk

Sinch Chatlayer

Dialogflow

Cognigy

Jobbeschreibung

AI Operations Specialist (Customer Support)

Department: Customer Operations / Operational Excellence

About Sinch

At Sinch, we are shaping how the world communicates. Our cloud communications platform reaches billions of mobile phones worldwide through SMS, messaging apps (WhatsApp, RCS, Messenger), voice, and email. We power meaningful customer conversations for the world's largest enterprises.

As part of our commitment to operational excellence and industry-leading customer experience, we are transforming how we support our clients using AI-driven automation, conversational intelligence, and scalable workflow design.

Key Focus Areas & Responsibilities

Map out intuitive, self-service conversation paths and resolution flows across messaging channels and web support, from first touch to resolution or smooth escalation.

Define and maintain consistent agent personas, tone of voice, empathy guidelines, and brand-aligned dialogue tailored to various customer tiers and technical complexity.

Architect conversational logic for ambiguous user intents, edge cases, and graceful failure handling.

AI Optimization & Prompt Engineering

Formulate, refine, and maintain high-performing system prompts, few-shot examples, and dynamic context parameters.

Work closely with knowledge bases and Retrieval-Augmented Generation (RAG) frameworks to ground AI answers in verified Sinch documentation.

Implement robust boundary rules, safety filters, and anti-hallucination guardrails to protect customer data and uphold brand integrity.

Build and run synthetic test suites, benchmark datasets, and regression checks to assess response accuracy, tone consistency, and task completion prior to production deployment.

Continuously review real-world conversation logs, agent handoffs, and customer drop-offs to pinpoint failure patterns and areas of friction.

Drive ongoing A/B testing on prompt variants, conversational hooks, and journey branches based on empirical CSAT, First Contact Resolution (FCR), and containment metrics.

Tooling Integration & Cross-Functional Alignment

Collaborate with Support Engineering and IT to link conversational AI agents with core ticketing platforms (e.g., Zendesk), CRM systems, and internal Sinch APIs.

Partner directly with front-line support managers and product teams to translate emerging product updates and user pain points into automated workflows.

What We Are Looking For

Core Competencies (Must-Haves)

Demonstrable experience structuring multi-turn dialogue trees, conversational UI flows, and tailored bot personas in an enterprise or SaaS setting.

Strong command of prompt engineering techniques, LLM behavior, context windows, and strategies to minimize hallucination while maximizing resolution accuracy.

Proven track record of using empirical data, conversation analytics, and structured evaluation rubrics to continuously benchmark and iterate on AI performance.

Deep understanding of support operational metrics (CSAT, deflection/containment rates, Average Handle Time, escalation quality).

Technical & Platform Experience

Hands-on experience with conversational AI platforms (e.g., Sinch Chatlayer, Google Dialogflow, Cognigy) or modern generative AI agent builders.

Familiarity with customer service ticketing systems (Zendesk, Jira Service Desk) and knowledge base management.

Basic understanding of APIs, webhooks, and JSON data structures to understand how conversational agents pass and receive data.

Mindset & Attributes

You let user interaction transcripts and performance data guide your design decisions.

You stay ahead of rapid shifts in generative AI capabilities and are eager to test new approaches.

Comfortable translating technical AI concepts into clear business logic for operational stakeholders.

Preferred Qualifications

Experience working in CPaaS, telecommunications, cloud software, or high-volume customer service operations.

Hands-on experience evaluating RAG architectures and embedding-based search retrieval systems.