Epiminds unveils an AI "marketing manager," raises $6.6M to build full-stack AI teams
Stockholm-based Epiminds launched with $6.6M in seed funding to ship a multi-agent AI system that acts like a complete marketing team. The company was founded by Elias Malm (ex-Google) and Mo Elkhidir (ex-Spotify) to shift focus from clicks to creativity.
The core product is Lucy, an AI marketing manager that directs 20+ specialized agents. They handle budget pacing, performance reporting, bidding strategies, and creative analysis-end to end. Agencies can onboard a client in under 30 seconds and spin up an AI-run campaign team.
Why this matters for marketers
Most teams are stuck in spreadsheets, dashboards, and manual optimizations. Epiminds targets that drag. As Elkhidir put it, marketers are being asked to do more with less, and the busywork is the bottleneck.
Their system connects insights directly to execution, learns your playbooks, and flags risks before they become problems. The aim isn't to replace strategists. It's to give them leverage.
How Lucy operates
- Orchestration: A manager-agent (Lucy) delegates to specialists across bidding, budgeting, creative analysis, and reporting.
- Speed: New client setup in under 30 seconds, then immediate task allocation.
- Learning: Adapts to an agency's processes and playbooks over time.
- Guardrails: Proactively flags anomalies and spend risks for human review.
Think of it as an AI workforce coordinating routine operations so your team can focus on positioning, offers, and creative direction.
Traction and funding
In 12 weeks, agencies managing 240+ brands have signed on. The round was led by Lightspeed Venture Partners, citing the ambition and scope of the multi-agent model.
What to do next if you lead marketing
- Audit your workflow: List repetitive tasks across channels (budget checks, QA, report builds, creative tagging). Prioritize what's safe to automate first.
- Start narrow: Pilot one channel or one account. Define clear success metrics: hours saved, error reduction, and lift in creative testing velocity.
- Set guardrails: Approval thresholds, budget caps, brand rules, and escalation paths. Keep humans in the loop for creative and strategic calls.
- Integrate data: Ensure clean UTM standards, CPM/CPC/CPA baselines, and conversion definitions so agents can make sensible decisions.
- Measure impact weekly: Compare time-on-ops vs. time-on-strategy, CAC/ROAS shifts, and the number of creative variants tested.
What to watch next
- Autonomy levels: Which tasks can safely move from "recommend" to "auto-execute" without risk to brand or budget.
- Creative analytics: How well the system links asset features to performance and suggests new concepts grounded in data.
- Reporting clarity: Transparent logs of actions taken by each agent and the rationale behind them.
- Agency economics: New pricing models as ops hours decline and strategic value rises.
Bottom line
Multi-agent AI is moving from tool to team. Epiminds is betting marketers want an always-on workforce that handles the grind, so humans can focus on strategy, offers, and creative craft.
Upskill your team: If you're building an AI-ready marketing org, consider the AI Certification for Marketing Specialists to standardize skills across channels and workflows.
Your membership also unlocks:
