PAM Core (LLM & Conversation Quality)

Company

Pam is an AI playground for builders. We combine strong engineering fundamentals with cutting-edge AI, creating space for you to grow into the engineer of tomorrow.

We're backed by Autotech Ventures and partner with the biggest names in automotive (Tekion, Dealer-FX, CDK). We've powered $100M+ in repair orders and handle calls for 500+ dealerships across the US and Canada.

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. When Pam says the wrong thing on a call, customers notice — and so do we.

This is support engineering done right — you'll ship code, improve AI systems, and own problems end-to-end.

We're hiring Support Engineers to own LLM and conversation quality issues. You'll debug why Pam said the wrong thing, fix it (code or prompts), and build systems so problems don't repeat. This isn't traditional support — you'll ship PRs, improve prompts, and build runbooks that make the whole team more effective.

If you want to understand AI systems deeply, ship real fixes, and own problems end-to-end — this is your role.

Pam.ai is not able to provide employment visa sponsorship at this time. Candidates must be authorized to work in the United States without current or future sponsorship.

What You'll Do

• Debug LLM Behavior: Analyze call transcripts to find where Pam went wrong — model issue, context issue, prompt issue, or config issue
• Handle Custom Setups: Work with unique customer configurations, phone system issues, and custom AI instructions
• Ship Fixes: Open PRs for bugs you find. Don't just diagnose — fix it, test it, deploy it
• Improve Prompts: Tune prompts to reduce hallucinations, failed intents, and bad responses
• Build Runbooks: Create documentation and tools so the ops team can handle more issues without escalating to you
• Protect Your Time: Push back when issues get escalated that shouldn't. Every repeat escalation is a failure — fix it permanently or build a runbook

Requirements:

• Have 1-3 years of software engineering experience
• Are strong in TypeScript (our entire stack) and AWS
• Use AI to code — Claude Code, Cursor, Copilot, or equivalent. This is how we work.
• Have strong debugging skills — you can read logs, trace issues, and find root cause
• Can ship PRs — you don't just find problems, you fix them
• Write clearly — you'll communicate with customers and build documentation
• Care about customers but also protect your time — you hate seeing the same issue twice
• Are curious about LLMs and want to understand how they work (and fail)

Prefrences:

• Experience with LLMs or prompt engineering
• Experience with voice AI or conversational systems
• Built internal tools or runbooks before
• SQL skills for querying customer data
• You've done support before and found ways to automate yourself out of repetitive work