AI Support Systems Engineer

About the Role

Pam is the fastest-growing voice AI platform for auto dealerships. We handle millions of calls, book service appointments, and drive real revenue for 500 dealerships.

The Outbound team is hiring a Support Engineer to own the issues that sit between customers, our dashboard, and our AI systems. You'll triage support tickets, reproduce dashboard bugs, debug LLM behavior, fix issues in code, config, or instructions, and build systems so the same problems do not repeat.

This is not traditional support. You'll use AI coding tools, ship PRs, improve LLM instructions and evals, and turn repeated support work into runbooks, automations, and product fixes.

The long-term mission is to make support an autonomous product surface: AI handles the repeatable triage and resolution work, humans are escalated only when judgment or customer context is truly needed, and every ticket makes the product smarter.

If you like debugging product behavior, tracing weird edge cases, using AI to move fast, and owning problems end-to-end, this is your role.

What You'll Do

• Triage Support Issues: Separate user confusion, dashboard bugs, product regressions, configuration issues, data problems, and LLM behavior failures.
• Debug Dashboard Bugs: Reproduce UI issues in the Outbound dashboard, inspect browser/devtool state, trace API responses, and write clear bug reports or fixes.
• Diagnose LLM Failures: Review prompts, instructions, tool calls, conversation traces, and customer outcomes to understand why the AI did the wrong thing.
• Ship Fixes: Open PRs for bugs you find. Do not just diagnose - fix it, test it, and get it shipped.
• Improve AI Instructions: Tighten LLM instructions, support macros, routing logic, and eval cases so support quality improves over time.
• Build Runbooks: Create documentation and tools so ops and customer-facing teams can handle more issues without escalating to engineering.
• Prevent Repeat Escalations: Every repeated issue should become a product fix, config fix, instruction update, automation, or runbook.
• Build the Support Engine: Help create autonomous triage, answer drafting, verification, routing, and escalation workflows so humans only handle the exceptions.
• Close the Loop with Product: Turn ticket patterns into clear product feedback, regression checks, and support automation opportunities.
• Work In Person: Collaborate from our Tysons Corner office 5 days/week so support, product, and engineering stay tightly connected.

You Might Be a Good Fit If You

• Have 1-3 years of software engineering, support engineering, QA engineering, or similarly technical experience.
• Are strong in TypeScript or can work comfortably in a TypeScript-heavy codebase.
• Use AI to code - Claude Code, Codex, Cursor, Copilot, or equivalent. This is how we work.
• Have strong debugging skills - you can read logs, inspect network requests, reproduce issues, and find root cause.
• Understand web app basics - frontend state, APIs, browser devtools, auth/session issues, and how UI regressions happen.
• Are interested in LLM product quality - prompts, instructions, hallucinations, tool failures, evals, and agent behavior.
• Can ship PRs - you do not just find problems; you fix them.
• Write clearly - you will communicate with customers, document issues, and build runbooks.
• Are patient with customers but protective of your own time.
• Want a role that starts in support and grows into owning support as a product surface.

Bonus Points

• Experience with SaaS dashboards, customer support tools, QA, or product operations.
• SQL skills for querying customer or product data.
• AWS or observability experience: CloudWatch, DynamoDB, S3, logs, traces, or similar tools.
• Experience with LLM evals, prompt/instruction writing, or AI agent testing.
• Built internal tools, support automations, or runbooks before.
• Worked in a fast-moving startup where support, product, and engineering overlap.

Compensation:

Base Equity