AI Support Engineer Lead (US)

Dust•San Francisco, CA

7d•$240,000 - $350,000•Hybrid

About The Position

The AI Support Engineer Lead at Dust will redefine the support function by applying an AI-first mindset. This role involves a dual focus: handling complex customer issues when automated systems fall short, and crucially, building the AI systems that progressively reduce this manual workload. Support at Dust is viewed as a product, with success measured by eliminating ticket categories rather than just resolving them. The individual will define, implement, and continuously improve the infrastructure for world-class support at scale, including AI agents, automation workflows, classification systems, knowledge systems, and tooling built on Dust itself. This role requires deep dogfooding of the product and is open to experienced support engineers ready to build systems at scale.

Requirements

Function-building experience: Defined or significantly shaped a support engineering function (operating model, standards, hiring, process rebuilding).
High technical aptitude: Comfortable reading code, analyzing logs, navigating codebases, and troubleshooting distributed systems.
AI fluency at a builder level: Proven track record of building custom agents, automations, or workflows using AI tools (e.g., Dust, Cursor, Claude Code, n8n).
Exceptional prioritization skills: Ability to quickly and confidently determine what to solve immediately, automate, escalate, or delegate.
People development instincts: Experience coaching and growing engineers, providing effective feedback.
Bidirectional communication skills: Ability to translate technical concepts to non-technical audiences and vice versa.
Interest in or experience with a Build/Run split: Viewing support as more than just ticket resolution.
Builder mentality: Instinctively automates repetitive work and holds the team to the same standard.
Resilience-empathy paradox: Ability to absorb customer frustration while caring about user experience.
Leads by example: Willingness to handle difficult tasks and be present in the support queue.
Thrives in ambiguity: Enjoys building systems from scratch in undefined environments.
Low ego, team player: Shares knowledge freely and prioritizes team success.
High ownership: Proactively identifies and addresses functional gaps.
Systems-first thinking: Focuses on eliminating categories of work, not just faster resolution.
Pattern recognition to repeatability: Turns recurring issues into scalable systems, automations, workflows, or documentation.
Deep investigation instincts: Investigates thoroughly before escalating, providing high-context information.

Nice To Haves

Even if you don’t meet every requirement, we encourage you to apply based on curiosity, potential, and determination.

Responsibilities

Set the technical direction for the support stack, determining automation priorities and standards.
Design, ship, and maintain AI agents and automation workflows for tasks like ticket classification, response drafting, and incident detection.
Identify recurring issue categories and engineer solutions through automation, documentation, prompt iteration, or product feedback.
Build and maintain tooling (e.g., MCP integrations, Dust agents, internal scripts) to enhance team capacity.
Define and manage the 'Support as a Product' backlog, including prioritization and execution coaching.
Hire, onboard, and develop support engineers.
Handle the most complex support tickets and set the standard for issue resolution.
Investigate complex issues using logs, code, and internal tools to find root causes.
Manage escalated cases with clear communication for diverse audiences.
Analyze agent responses to identify inconsistencies and iterate on prompts, documentation, and tooling to minimize human intervention.
Represent support in engineering and product discussions, synthesizing and prioritizing feedback.
Build relationships with engineering and customer-facing teams for efficient escalations.
Own the end-to-end feedback loop, closing gaps in information across engineering, product, and documentation.