Own every conversation the AI agent cannot resolve, and resolve it. Most of what reaches you should end with you. Escalate only when the answer genuinely sits elsewhere: to the Product Owner as a clarification or feature request, to Engineering as a bug with full reproduction steps, or to the right contact on the sales team when the inquiry is commercial rather than technical. Knowing the difference between a question you should work harder on and one that genuinely belongs to someone else is the core judgment of this job.
Improve the resolution rate of our Intercom AI support agent. Review the conversations it handled poorly or handed off, correct or write the underlying content, and track the share of conversations it closes without a human. Every question it learns to answer is one that never reaches the queue.
Hold a first response within 30 minutes on email and Intercom during your shift. Every conversation that reaches you has already been through an AI handoff, so the wait the customer feels starts with you.
Troubleshoot customer API integrations: authentication and API keys, request and response payloads, HTTP status and error codes, rate limits, pagination, async and bulk job workflows, and sandbox versus production behavior.
Support customers who consume data through file-based delivery in addition to the API, including file format, schema, field definition, and record count questions.
Reproduce and document defects. Capture the request, response, identifiers, and timestamps, and open well-formed Jira tickets for Engineering. Track each one to closure and keep the customer updated along the way.
Join occasional follow-up calls and screen shares when a written exchange is not resolving the issue fast enough.
Work with the Product Owner to write and maintain help center articles, draft product release notes, and keep our API documentation accurate as endpoints and fields change.
Spot recurring issues and bring the themes back to Product and Engineering.
4+ years in technical support, support engineering, or developer support for an API-first or data product, with a significant share of that time spent helping customers build and debug integrations.
Hands-on experience supporting a data product delivered through more than one channel, for example a REST API alongside bulk file or feed delivery.
Fluency in API concepts: REST, HTTP methods and status codes, JSON, authentication tokens, pagination, rate limiting, and webhooks. Comfortable reproducing a customer's request in Postman or curl.
Working knowledge of databases and query concepts. Enough to reason about schemas, field definitions, record matching and deduplication, and to read or write basic SQL.
Excellent written English, and clear, easily understood spoken English. Calls are occasional rather than daily, but you should be comfortable getting on a screen share with a US developer when writing is not moving things along.
Experience using an issue tracker, Jira preferred, to escalate and track defects with an engineering team.
Hands-on experience with Intercom, Zendesk, or a comparable support platform.
Demonstrated experience writing customer-facing technical content: help center articles, FAQs, release notes, or API reference documentation.
Experience supporting US-based customers from an offshore location, on a shift aligned to US business hours.
Experience curating the knowledge base behind an AI support agent (Intercom Fin, Zendesk AI, or similar) and measurably improving its resolution rate. Preferred.
Familiarity with OpenAPI/Swagger and a documentation platform such as Stoplight, Readme, or Redocly. Preferred.
Background in real estate, proptech, mortgage, title, or marketing data. Preferred.
Light scripting in Python or JavaScript to reproduce integration issues. Preferred.
