Support Engineer (AI-First)

Hedra•San Francisco, CA

About The Position

As a Support Engineer (AI-First), you will design, build, and run the systems that power Hedra’s AI-first customer support. You'll operate production-grade AI support workflows (routing, escalation, knowledge, and evaluation), and you’ll jump in as the human escalation point for key accounts when issues are complex, time-sensitive, or high stakes. You’ll partner closely with Engineering and Product to turn support signals into fixes, reliability improvements, and a better customer experience.

Requirements

Built or operated technical support for a developer-facing or complex product (B2B SaaS, infra, AI, APIs), including escalation ownership
Strong debugging skills across APIs + web apps (auth, logs/telemetry, repro steps) and ability to communicate crisp root-cause hypotheses
Comfortable reading and shipping code/scripts (Python and/or TypeScript) to automate workflows, fix sharp edges, or improve tooling
Experience building durable support operations: SLAs, severity definitions, incident comms, postmortems, and clear internal handoffs
Hands-on experience improving LLM-powered workflows (e.g., RAG, tool calling, prompt/tool iteration) with a focus on correctness and safety
Strong customer-facing communication and able to lead updates for key accounts during incidents and high-urgency escalations
Startup or 0→1 environment experience

Nice To Haves

Built evaluation/QA for AI systems (sampling, automated checks, regression tests, escalation-quality metrics)
Experience with Intercom/Zendesk/Front + knowledge base workflows (docs → KB → retrieval)
Background in observability/monitoring (Sentry/Datadog/OpenTelemetry) and incident response practices
Experience supporting generative media / video / creative tooling, or similar “high-variance” AI outputs
Experience managing vendors/BPO or building a small support team over time