This is a high-ownership, hands-on role at the intersection of people management and AI-driven automation in a lean, fast-scaling eCommerce business with multiple global brands.
You will own the customer support function end to end, managing 30–40 agents and team leads, and ensuring quality across 20,000+ monthly tickets while driving the next stage of AI automation.
70–80% of tickets are already handled by a custom LLM system. In this role, you will improve it, extend it into AI phone support, and turn CS into a real-time feedback loop for the business.
WHO THEY ARE
What started as a single air purifier sold during COVID has grown into a multi-brand eCommerce operation spanning supplements, home tech, personal care, and everyday gadgets, with 15 markets, hundreds of SKUs, and a small core team punching well above its weight.
What sets this company apart:
- Flat structure, generalist mindset: Most team members own multiple functions, not just one lane.
- Output-driven, never micromanaged: Results matter, not hours logged.
- Genuinely obsessed with AI: Not as a buzzword, but as a real operational edge they're building every week.
- Direct, feedback-rich culture: Problems get named and solved, no politics.
This is a fully remote, full-time contractor role.
WHAT'S IN IT FOR YOU
- Ownership: Run CS as a function, not a task list, with direct access to the CEO and Ops Manager.
- AI tooling: Experiment freely, test new tools, and build systems the way you think they should work.
- Compensation: Performance bonuses tied to refund rate, chargeback rate, and quality improvements.
- Growth: Clear career trajectory with room to grow.
- Autonomy: No micromanagement, you are measured on outcomes, not check-ins or hours.
- Team: Work with a fast-moving, high-performance team that moves quickly and executes seriously.
- Setup: Fully remote with flexible working arrangements.
WHAT YOU'RE GETTING INTO
- Own end-to-end customer support, including KPIs, performance, quality, response times, and AI across 20,000+ monthly tickets.
- Manage around 15 agents, ensuring SLA delivery through hiring, scheduling, capacity planning, and performance management.
- Improve LLM/Claude integration by refining prompts, fixing issues, shipping updates, and reducing 20–30% manual ticket volume.
- Build and launch AI phone support as a core roadmap initiative.
- Turn CS data into business insights by identifying ticket patterns and routing them to marketing, ops, and product.
- Reduce refund rate, chargeback rate, and reshipment rate with measurable P&L impact.
- Align CS execution with Ops Manager and CEO to support scaling across new and existing brands.
- Build a system of continuous improvement that replaces firefighting with compounding operational gains.
WHO YOU ARE
- You have more than 5 years of experience as a Customer Support Manager managing 15+ agents via team leads and owning hiring, onboarding, and performance.
- You have built and scaled CS teams, including processes, team lead management, and SLA delivery at volume.
- You have built or worked closely with AI/LLM systems in CS, ideally automating 70%+ of ticket volume with tools such as ChatGPT, Claude etc.
- You have experience in fast-paced direct response eCommerce, ideally with physical products and smaller to mid-size companies.
- You are comfortable with CS tools like Gorgias (the main customer support that we currently), Zendesk, Front, or Intercom, without being tied to a specific platform.
- You combine people management with systems thinking, moving between operations and AI-driven improvement.
WHAT SUCCESS LOOKS LIKE
-
30 days: Baseline key metrics (refunds, chargebacks, CSAT, AI automation). Ship initial AI improvements. Align with team leads through 1:1s. Full CS visibility established.
-
60 days: Improve AI performance through iterations. Start reducing the refund rate (target 10–15% trend). Launch feedback loop with marketing and ops.
-
90 days: Sustain 10–15% refund reduction. Advance AI into phone support (pilot or live). Turn CS insights into measurable impact across the business.