Operational intelligence Hosted in France and the EU
AI handles the interactions. Your teams decide.
Repeated questions, incomplete tickets: your project teams end up doing support. Weloop takes it on, inside your tools.
92% of requests resolved without your team.9 h given back each week to the people who handled support.
Trusted by
100,000+ users satisfied across Europe
Weloop answers your users, inside your applications. You only receive what needs a decision.
A request comes in. The AI resolves it, or you decide.
Weloop Assistant<script async src="https://js.weloop.ai/v1/<your-project-id>/widget.js"></script>- Access to Weloop
- A script in your applications, Teams or Slack
- 1 day to go live
Weloop in numbers
Everything is included, in one product.
Where your users already work
Built into your applications, in Teams or Slack, in 17 languages.
From your articles, and it shows you
Import PDF, Word, PowerPoint, Excel. Every answer links to its source.
Amounts are written with a dot, not a comma: 12.50.
Into your tools, as a complete case file
Jira, ServiceNow, Slack, Teams or the Weloop kanban, synced both ways.
You see everything. And every low score is followed up.
Satisfaction surveys in the assistant, by email or in your pages, followed through to the answer.
A low score triggers a follow-up
Everything is measured live
Everyone gains from the first week.
No more sorting. Time for decisions.
The time spent sorting goes back to what truly needs your judgment.
A real answer. Not a ticket lost in the void.
A sourced, instant answer, and nothing new to learn.
Proof in real time. Data in France and the EU.
Autonomy, SLA and satisfaction, visible live.
The calculation, with your numbers.
A cautious estimate. The real gain is then measured in your dashboard.
Show the calculationHide the calculation
Starting example: an application with 1,600 users. Without the number of requests, we count 2 requests per user per year, so 3,200. We only count 67% of requests as resolved by the agent, a cautious assumption, below the 92% measured by Weloop: 2,144 requests. Each one avoids a €20 ticket, a common order of magnitude for a level 1 ticket, to replace with your own cost: €42,900 per year. For the team, each avoided ticket frees 13.7 min, the average handling time of a written ticket (MetricNet, 2023): 490 h per year, or about 9 h per week. These hours are the same gain as the euros, seen another way: they do not add to them. For several applications, the calculation assumes they are the same size.
A European AI you can present to your CIO and your DPO.
Sovereignty and security in detailWhere is our data?
In France, with Scaleway, with backups in a second EU region. 99.5% availability SLA.
Which AI models?
Mistral AI, hosted in the EU, with no training on your data. No AI processing outside the EU. US integrations, such as Slack, remain optional.
Where can it be installed?
In the cloud, on your own domain or on your infrastructure (Kubernetes, AWS).
What about compliance?
GDPR by design, DPA on request, DPO registered with the CNIL, SAML / OIDC SSO, annual independent audit, breaches notified within 48 h.
It reads the page the user is on, answers from your knowledge base while showing its sources, and recognizes duplicates. It then asks the questions set out in your rules. At any time, the user can ask for your team to take over.
You describe your rules in natural language. Weloop creates the ticket in Jira, ServiceNow, Slack, Microsoft Teams or its built-in kanban, with statuses and replies synced both ways. For any other tool, it publishes the event to Amazon EventBridge.
Yes. DPA on request, a public list of subprocessors, export and erasure of a person's data within 10 business days, breaches notified within 48 hours. Raw IP addresses are never stored, and session replay requires consent.
Yes. For software you cannot modify, the browser extension (Chrome, Edge, Firefox) shows the same assistant on the sites you declare. It is deployed through your browser policy.
Yes. A REST API and an MCP server cover projects, tickets, announcements, surveys and the knowledge base. You can publish a release note from your pipeline, or let your AI agents query Weloop.
Pricing is set with our team, based on how you deploy: Weloop cloud, on your own domain or self-hosted. Contact us to discuss it. The three options are detailed on the Deployment options page.
See it on your use case.
- 1Demo. 30 minutes with our team, on your real use case.
- 2Getting started. Access to Weloop and a script, in a first application. Nothing changes for your teams.
- 3Deployment. On your other applications, in Teams and in Slack.
