Head of AI Sales

Groupon•Chicago, IL

1d•$200,000 - $300,000•Hybrid

About The Position

Groupon is rebuilding how it sells. This role is the Foundry bet for Supply — the most strategically important AI build in the commercial organisation. Reporting directly to the CSO, the Head of AI Sales builds and runs a production-grade fleet of AI agents across two layers. The first layer operates autonomously in parallel to the human sales team — running full sales motions independently: prospecting, outreach, qualification, reactivation, and inbound triage. The second layer augments human reps at the point agents reach their limits: when a rep steps in, they arrive fully equipped — account context surfaced, deal history logged, next steps prepared — and the agents continue working in the background. The goal is a 10x sales organisation: reps move from player to coach, focusing on judgment, relationships, and decisions while agents handle execution. What makes this build uniquely defensible is the data: proprietary consumer transaction signals, merchant health intelligence, and three years of behavioural sales call data that no competitor can replicate. You own both layers, the handoff between them, and the architecture that makes the whole system run. This role does not duplicate what existing functions own — it gives Groupon a parallel sales force, and makes the existing one sharper. North Star A two-layer AI sales system: agents that run full motions autonomously in parallel to the human team, and agents that make human reps sharper the moment they step in. Reps close deals and manage relationships. Agents do everything else — before, during, and after. And the system gets better over time: every signal, every call, every conversion feeds back into the fleet. Scale compounds. Investment does not.

Requirements

Top-tier academic background or equivalent formation in a high-calibre AI or technology environment — keeping the door open to exceptional self-starters who came up through companies like OpenAI, Palantir, or a serious AI startup rather than a traditional degree path.
7+ years at the intersection of AI and commercial/sales — enterprise AI SaaS, GenAI, or AI solutions. Builder track record required.
Shipped multi-agent sales automation in production: sequencing, routing, CRM automation, AI dialers, call analysis, or reactivation.
Full-stack fluency at Level 8+ agentic maturity: orchestrator + specialist agent architecture, parallel sessions, background cron agents, autonomous loops. LLM APIs (OpenAI, Anthropic), LangChain/Dust/CrewAI, Clay/Apollo/ZoomInfo, Python, Salesforce Opportunity-level flows.
Deployed AI SDR platforms (11x.ai, Artisan, Amplemarket, or equivalent) — you know where they break and have a clear build-vs-buy framework.
Commercial judgment across sales motions — you know what an agent can own end-to-end vs. where a human must take over, and you design the handoff intelligently. Every motion tied to a measurable ROI proof point.

Responsibilities

Build and own the agent architecture — Design the full orchestration layer: lead scraping and ICP-scoring, BD outbound sequencing, IB triage and routing, MD reactivation, and multi-touch National/Enterprise engagement — all running in parallel, production-grade from day one.
Run and iterate the fleet on live data — Own in-production performance. Define signal logic, trigger conditions, prompt architecture, and account prioritisation intelligence (PDS score, pyramid, campaign tags, open CRM cases).
Build and operate the call intelligence layer: process call transcripts at scale, score performance, extract competitive patterns, and surface category playbooks and coaching signals to reps and managers. Test on live merchant data, kill what doesn’t convert, ship before you’re comfortable.
Own both layers — autonomous motions and human augmentation — Layer 1: agents run complete workflows without rep involvement — prospect identification, outreach sequences, lead qualification, reactivation, and inbound first response. Reps receive warm, qualified opportunities, not cold lists. Layer 2: where an agent reaches its limit and a human steps in, the rep arrives equipped — account context surfaced, deal history logged, suggested approach ready. Agents continue to work in the background while the rep handles what requires judgment. You design both layers and the handoff between them.
Coordinate — don’t duplicate — Align with internal stakeholders across Sales, Commercial, and LOB leadership on what the fleet should solve next. You set the AI architecture; they deliver within it. Every request from the commercial organisation becomes a roadmap decision, not a distraction.
Own the vendor stack — Evaluate, deploy, and exit. Know where 11x.ai, Artisan, Clay, Amplemarket, and Dust break at scale. Clear build-vs-buy framework, no expensive lock-in.