Quack Raises $7M to Solve Customer Support Issues Before They Happen

Quack Raises $7M to Build Proactive, Agentic AI for Customer Support

Quack has raised a $7 million seed round led by Hanaco Ventures and Storytime Capital, with participation from Fusion VC, Savyon Ventures, Seed IL, and private investors including WalkMe CEO Dan Adika. The capital backs U.S. expansion and a clear shift in mindset: move support from reactive ticket-taking to proactive issue prevention.

Why Reactive Support Is Breaking

Most teams operate in backlog mode. Tickets pile up, costs climb, and key issues never get reported. The result is silent churn and stressed agents who spend their time firefighting instead of creating value.

What Makes Quack Different

Quack runs an agentic AI operating system that takes initiative. It learns your product and customer base by topics (similar to onboarding a new specialist), monitors patterns across data sources, and detects early signals of friction.

When it can fix the issue, it acts. When it needs a human, it escalates with full context so customers don't repeat themselves and agents start from a running start.

Every resolved issue feeds a quality loop. Both human agents and AI agents get smarter with each pass, shrinking the volume that ever becomes "support."

What This Means for Support Leaders

• Fewer repetitive tickets; more time for complex work and relationship building.
• Fewer surprise escalations; more predictability in staffing and SLAs.
• Higher CSAT through prevention, not apologies.
• A clearer line from support to retention and expansion.

High-Impact Use Cases You Can Deploy

• Billing hygiene: correct invoice errors before sending; flag unusual charges.
• Onboarding friction: detect stalled adoption and trigger in-product guidance or outreach.
• Service health: spot interruptions or degraded features and auto-communicate workarounds.
• Entitlements and access: catch misconfigurations before they block workflows.
• Renewal risk: surface early churn signals from usage, sentiment, or unpaid invoices.

How Agentic AI Works (In Practice)

• Topic-based training: teach the AI your product areas, policies, and user journeys.
• Signal detection: monitor logs, product analytics, billing, and comms for early warning signs.
• Autonomous actions: fix what's safe to fix; communicate status proactively.
• Smart escalation: hand off to humans with history, steps taken, and recommended next moves.
• Continuous learning: fold outcomes back into the system to reduce repeats.

For a primer on AI agents and why initiation matters, see this overview from NVIDIA: What is an AI agent?

90-Day Rollout Plan

• Week 1-2: Map your top 5 friction themes (billing, onboarding, outages, access, renewals).
• Week 3-4: Connect data sources (product analytics, CRM, billing, status logs, support history).
• Week 5-6: Define safe actions and escalation rules; set thresholds to avoid noise.
• Week 7-8: Pilot with one product area and a small customer segment.
• Week 9-10: Measure impact; expand to the next two themes.
• Week 11-12: Formalize playbooks and governance; train the broader team.

Measure What Matters

• Prevented tickets per week (issues resolved before contact).
• Time-to-detection and time-to-resolution for proactive cases.
• First contact resolution and cost-to-serve deltas.
• CSAT for proactive communication vs. reactive tickets.
• Churn deflection and expansion triggered by proactive saves.

Risks and Guardrails

• False positives: set thresholds and require confirmation for sensitive actions.
• Privacy: log every action, limit data access, and respect consent preferences.
• Human-in-the-loop: keep humans as final approvers for high-impact changes.
• Change management: show agents the "why," give them context-rich tools, and reward prevention.

Why the Funding Matters

$7M gives Quack the runway to expand in the U.S., refine topic-based training, and strengthen integrations across the support stack. Backing from Hanaco Ventures, Storytime Capital, and operators like WalkMe's CEO signals confidence in proactive support as a core function, not a side project.

Where This Is Heading

Analysts expect agentic AI to handle the bulk of common support tasks by the end of the decade. The scorecard won't be how fast you reply; it will be how few problems customers experience in the first place.

Support becomes an invisible layer of the product. Quiet, consistent, and measurable.

Next Steps

• Skill up your team on AI, automation, and customer experience ops: AI courses by job role.
• Study agent frameworks and decide where autonomous action is safe vs. where human approval is required.
• Start small, measure, then expand. Prevention compounds.