On the radar — not actively interviewing. This role sits on our 12-month hiring roadmap rather than today's pipeline. Submit the talent pool form to be considered when we begin interviews.

Technical Support Engineer — Cantonese (Hong Kong / Guangdong)

○ On the radar · talent pool onlyRemoteFULL_TIME / CONTRACTORFocus region: Anywhere

About the role

DogeSMS serves customers in 60+ countries; first-line technical support runs as a follow-the-sun rotation across language regions. This radar role is the Cantonese (Hong Kong / Guangdong)-speaking seat in that rotation, focused on Hong Kong, Macau, and Cantonese-speaking communities in Guangdong and the Pearl River Delta. Hong Kong's fintech and crypto exchange ecosystem (under HKMA and SFC oversight) generates intensive A2P SMS volumes that customers expect to receive in Cantonese-localized templates. Cantonese is intentionally a separate language seat from Mandarin — the markets, the regulatory context, and the customer phrasing are different enough that pooling them degrades support quality. The role is intentionally different from a typical SaaS support seat. Roughly 60% of weekly hours go to working alongside our customer-facing AI agent: reviewing transcripts in Cantonese (Hong Kong / Guangdong), labelling edge cases for retraining, and partnering with the AI Agent Platform team on evaluation datasets that capture region-specific phrasing ('驗證碼', 'OTP 短訊', '即時付款', '收唔到短訊' (when a customer says they didn't receive)). The other 40% is direct human support for tickets the agent escalates — integration debugging with customer engineers, payment reconciliation questions, carrier-side delivery investigations. Hong Kong / Guangdong covers UTC+8 with peak hours overlapping Mandarin and Japanese seats; Cantonese-specific evening hours (when HK fintech support volume peaks) get dedicated coverage.

What you'll do

  • Own first- and second-line technical support for Cantonese (Hong Kong / Guangdong)-speaking customers — primarily async (Intercom / email), with scheduled video sessions for enterprise accounts. Tickets routinely span carrier delivery, API integration, and payment reconciliation.
  • Serve as the language-specific training data partner for our customer-facing AI agent: review Cantonese (Hong Kong / Guangdong) agent transcripts weekly, label correctness, build evaluation datasets that cover regional phrasing nuances ('驗證碼', 'OTP 短訊', '即時付款', '收唔到短訊' (when a customer says they didn't receive)).
  • Translate region-specific technical context for English-language internal hand-offs — when a Cantonese (Hong Kong / Guangdong) customer reports a carrier-side delivery issue, the engineer on the other end of the escalation needs the technical detail in English without losing nuance.
  • Maintain Cantonese (Hong Kong / Guangdong) versions of customer-facing documentation, status page messaging, and incident post-mortems where customer impact warrants region-specific communication.

What we expect

  • Native or near-native Cantonese (Hong Kong / Guangdong) (written and spoken) with technical English at C1+ level — the AI agent training work and internal escalations require fluent English documentation.
  • 2+ years in technical support or technical account management at a developer-facing SaaS, ideally in payments, telecom, or identity / authentication tooling. Experience reading API responses and explaining HTTP error codes to customer engineers.
  • Direct prior experience working alongside AI agents in a support context — you have audited LLM responses, built eval cases, or partnered with an ML team on improving customer-facing agent quality. This is not a "willingness to learn" requirement; it is the primary differentiator for this role.
  • Comfort with the on-call rhythm of distributed-team support: timezone-overlapping with neighboring language seats, English hand-offs at shift change, no expectation of office attendance.

Compensation

On the radar. Compensation is calibrated to local cost of living when we open this role — we publish bands when we begin interviewing, not before. USD-denominated regardless of jurisdiction, contractor or employment depending on your legal situation.

Hiring process

  1. Not actively interviewing right now — this role sits on our 12-month hiring roadmap rather than today's pipeline.
  2. Submit the talent pool form below; we reach out when we begin interviews, typically 4-8 weeks before a public hire.
  3. When the role opens, expect a process similar to our current active openings: a short async exercise, 1-2 conversations with the relevant team, and a scoped paid trial project. We do not run leetcode-style screens.

Join the talent pool

Submit the form below to be considered when this role opens. We typically reach out 4-8 weeks before we begin interviews.

Email careers@dogesms.com →