AI Program Manager
Bloomreach builds agentic AI systems for our customers. We also run them ourselves.
Our AI SDR program books sales meetings with AI - inbound and outbound, across four regions, in multiple languages. The program needs an owner: someone who sets its direction, drives its results, and proves its ROI. This is a program leadership role, not an engineering one. Two dedicated AI ops experts build and run the platform. You decide what it does, why, and for whom. You should understand AI agents well enough to direct the people building them - your leverage is judgment, but being tech savvy is a must.
The team experiments by default: we test new AI tools, run scoped POCs, and keep what works. You'll own the program, its direction, its results, its pipeline number.
What you'll own
The AI SDR program, end to end. Inbound and outbound, AMER + EMEA + APAC. Who gets enrolled, what the agents say, when a human takes over, and what it produces in pipeline. You set direction; the platform ops managers make it run. (Inbound request ops and routing sit with a colleague.)
Pipeline as your number, ROI as your story. Meetings booked → accepted → qualified pipeline created. Cost per meeting, cost per pipeline dollar. You report the program's return to leadership, with numbers that hold up under questioning.
The human SDR side. SDRs are the other half of the program - and where it most often breaks: lead quality disputes, unclear prioritization, follow-up that dies after one touch, handovers that lose context. You'll sit with the SDR teams and their leaders, find root causes, and fix them - with an AI assist, a process change, or a better definition.
The human–agent split. What agents do, what humans do, and where the handover sits. You own that line and move it based on evidence.
Signal-driven enrollment. Targeting runs on manual batches today. You'll drive the move to continuous enrollment based on account scoring, intent signals, and engagement decay. You define the logic; the ops team builds it.
The buying experience. Outreach should arrive because an account signalled something, say something true about the buyer's business, and make the next step obvious.
Program reporting. Volume, coverage, results, and ROI on a live dashboard, generated automatically. Parts still run on exports. Fix that early.
What comes next. The AI SDR pattern - signal in, agent acts, human closes - applies beyond outbound: event follow-up, hand-raiser nurture, re-engagement. Once the core program runs well, you'll make the case for the next one and drive its launch.
What success looks like after 3 months
Outbound volume and coverage are consistent across regions. Nobody asks "why did outbound drop last week?"
Leadership references your ROI model in planning: cost per meeting, cost per pipeline dollar, trend.
The top three SDR pain points from your first quarter are root-caused and fixed - and SDR leaders would confirm it unprompted.
Meeting acceptance is high. Rejections have known causes, not theories.
Reporting is fully automated. Your time goes into decisions.
The program activities and volume is connected to human activities, pipeline number for ICP accounts is bigger - and you can show which changes drove it.
At least one new agentic program is live and producing pipeline.
What you'll bring
You've run programs with a number attached. Demand gen, SDR/BDR, growth, or GTM programs where you owned the outcome and defended the metric.
You've operated AI workflows in production. You've deployed, directed, or owned automations other people depended on. You can set requirements, challenge the build, and tell when the agent is the problem versus the data. You don't need to write code - you do need to never be bluffed by it.
ROI fluency. You can build a cost-per-outcome model and explain it. You'd rather show a live dashboard than a slide.
You diagnose team problems at the source. "The leads are bad" is where your investigation starts, not where it ends. You've worked with frontline teams, found the real constraint, and fixed it.
SDRs, AEs, and their leaders are your users. When the program misfires, you’ll be their point of contact.
The buyer's point of view. Buyers don't care that outreach was efficient - only that it was worth their time. You'll kill a volume that doesn't clear that bar.
Enough technical depth to be dangerous. You can read a workflow, question a data model, and learn a new tool in a week. Experience with our stack is welcome, not required.
Global comfort. Four regions, several languages, overlap with US hours.
Big Advantage:: hands-on LLM tooling beyond vendor UIs (Claude Code, MCP, custom agents); time in or working with SDR/BDR and demand gen function in B2B SaaS.
The stack you'll drive
Salesforce · Qualified · Salesloft · UserGems · Claude - plus a mandate to evaluate AI-native tooling with a clear bar for measurable impact.
The team
You'll join Marketing Operations six people who operationalize and automate everything in marketing. Two of them work full-time on the AI SDR platform: they build and run it; you set its direction and co-own it: reporting and attribution, co-own and maintain inbound/outbound processes and find new tools. It's a team that builds its own tools rather than filing tickets for them.