Davos 2026 gets EVA, Salesforce's AI concierge in the WEF app

Davos is giving every attendee an AI concierge. Salesforce's EVA cuts decision overload, curates sessions, updates plans in real time-a playbook event teams can pilot fast.

Published on: Jan 20, 2026
Davos 2026 gets EVA, Salesforce's AI concierge in the WEF app

Davos brings an AI concierge to every attendee - here's what event teams can copy now

The World Economic Forum is putting an AI "concierge" in every attendee's pocket. Salesforce built EVA (Event Virtual Agent) into the official Davos app to help guests cut through 450 sessions, shifting plans, and nonstop meetings.

The problem it solves isn't walking from venue to venue. It's decision overload. EVA trims prep time to minutes, flags what matters for each person, and keeps schedules coherent as the day changes.

What EVA actually does

  • Creates instant, context-aware briefs for sessions based on attendee interests and role.
  • Recommends next-best actions: which session to attend now, who to meet, where to spend limited time.
  • Adjusts plans as changes hit - so attention stays on content and people, not logistics.
  • Reasons over WEF's trusted data with permissions and governance intact.

Under the hood, EVA runs on Salesforce's Agentforce platform and borrows from the company's Slack-based personal assistant work. WEF is also rolling out AI agents to support staff during planning and live operations.

The stack behind the scenes

Salesforce's forward-deployed engineering team is implementing the concierge across the Davos app and wider Forum infrastructure. MuleSoft connects the Forum's CRM to finance, HR, travel, and operations systems, while Tableau tracks engagement, participation, and account activity across initiatives.

For event leaders, that mix matters: data unification for context, an agent to turn context into actions, and analytics to prove what worked. You don't need Davos scale to apply the same pattern.

See Salesforce Agentforce

Why this matters for hospitality and events

  • Reduces attendee confusion without adding more staff on-site.
  • Personalizes content and meeting discovery so rooms fill smarter, not just fuller.
  • Supports capacity management by nudging traffic away from bottlenecks.
  • Improves sponsor ROI with better-fit introductions and timely prompts.
  • Boosts accessibility with real-time summaries and simple next steps.

How to pilot an AI concierge in 90 days

  • Scope: Pick 1-2 attendee journeys (e.g., session selection + meeting scheduling).
  • Data audit: Map what the agent can read (agenda metadata, rooms, capacity, profiles) and what it must not.
  • Permissions: Enforce role- and account-based access. Keep PII in its lane.
  • Agent behaviors: Define approved actions (briefing, recommending, rebooking) and escalation rules.
  • Integrations: Start with CRM, agenda, rooms/capacity, push notifications, and helpdesk.
  • UX: Ship inside your existing app with clear "Ask EVA" entry points and simple prompts.
  • Training: Prep your concierge team to review and adjust agent suggestions in real time.
  • Contingency: Fallback playbook for outages: static agendas, SMS alerts, staffed info points.

KPIs to watch

  • Session fill rates vs. capacity and overflow incidents.
  • Helpdesk ticket volume and first-response time.
  • Change-communication latency (minutes from update to attendee awareness).
  • Meeting conversions and sponsor-qualified leads.
  • Attendee CSAT/NPS and qualitative comments about clarity and flow.

Risks and how to handle them

  • Wrong answers: Ground the agent in approved content with retrieval, and log citations.
  • Privacy: Enforce least-privilege access; separate public info from sensitive records.
  • Bias: Audit recommendations by role, company size, geography, and accessibility needs.
  • Over-automation: Keep a human-in-the-loop for reassignments, VIP cases, and emergencies.
  • Accessibility: Support concise text, screen readers, and language options.

What leaders are saying

Salesforce's CEO, Marc Benioff, says EVA goes beyond chat to act as a data-grounded concierge that helps people move faster, prepare better, and take action. WEF's president and CEO, BΓΈrge Brende, notes the Forum is applying agents to staff workflows and putting its institutional knowledge directly in attendees' hands.

Bottom line for event pros

Use Davos as a model: connect your data, define a small set of high-value agent actions, prove impact with analytics, and expand. The payoff is fewer frantic questions, better room utilization, happier sponsors, and attendees who feel looked after - without adding headcount.

If your team is ready to skill up for this shift, explore practical AI paths by role here: Complete AI Training - Courses by Job


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)
Advertisement
Stream Watch Guide