Key Takeaways
- World2Meet’s Intelligent Process Automation (IPA) initiative has saved the company over 33,000 operational hours annually, with measurable gains in contact centre efficiency and accuracy.
- The “invisible AI” approach embeds automation into existing workflows — email classification, sentiment analysis, corporate mobility — without customer-facing exposure.
- W2M’s results show that back-office AI integration, not just customer-facing tools, can drive significant productivity gains and build broader organisational appetite for automation. World2Meet saved over 33,000 operational hours in a year — not by launching a flashy AI product, but by quietly embedding automation into the workflows employees use every day. The travel division of Iberostar Group calls it “invisible AI,” and the results are hard to argue with. Their IPA initiative, a finalist at the CIO 100 Awards, is a solid example of what back-office automation actually looks like when it ships.
World2Meet Embraces Intelligent Process Automation
W2M launched its Intelligent Process Automation initiative in 2023 with a straightforward brief: find the processes worth automating and make them faster. By 2025, the results were tangible. The approach — embedding AI into existing systems rather than building standalone tools — is what W2M’s CIO Joan Barceló describes as the real driver of value. It runs behind the scenes, touching workflows that employees interact with daily but that customers never see.
Transforming Contact Center Operations
The contact centre is where the numbers get interesting. In January 2026 alone, the system automatically processed around 165,000 emails, classifying them with roughly 92% accuracy and running sentiment analysis on each one. In a large share of cases, it also pulled in reservation details to enrich the data before a human ever touched the email.
The practical effect: agents receive emails that are already classified, summarised, and loaded with relevant context. That pre-processing is where the 33,000 annual hours of savings come from. Incoming calls get the same treatment — transcribed and analysed so agents start each interaction already up to speed. If you’re thinking about how to evaluate whether a setup like this is actually working in production, the PARE framework for proactive agent evaluation is worth a look.
Streamlining Corporate Mobility and Third-Party Integrations
The automation extends into corporate mobility, where business travel requests typically arrive as unstructured emails or messages. The AI parses the incoming text, searches across systems for transport and accommodation options, and assembles a draft proposal. A human manager reviews and edits before anything goes to the client — which is the right call. Speed gains without a human checkpoint on outbound proposals is a risk most operators shouldn’t take.
W2M has also connected the IPA system to third-party platforms. One example is an integration with Samsara, used by W2M’s carrier operation in Mexico. The connection adds intelligent monitoring with geolocation and camera feeds, and according to W2M has contributed to fewer speed-related incidents, lower fuel costs, and reduced maintenance spend — a good illustration of how agentic integrations can extend beyond internal systems entirely.
Beyond Time Savings: Enhanced Quality and Consistency
Barceló’s framing is worth noting here: the headline isn’t just hours saved, it’s consistency. Automated classification doesn’t have bad days. It applies the same logic to the 165,000th email as it did to the first. W2M points to improved accuracy, fewer errors, and stronger alignment with corporate policies as outcomes that compound over time in ways that raw time metrics don’t fully capture.
This internal IPA work also sits separately from W2M’s customer-facing AI. Their generative AI assistant Mía, shown at FITUR in 2024 and 2025, handles direct customer interaction. The invisible AI layer is a different beast — infrastructure-level automation that makes the whole operation run cleaner, whether Mía is involved or not.
Fostering a Culture of AI Adoption
One of the less-obvious outcomes from W2M’s rollout is what happened internally. IT led the early use-case discovery, as you’d expect. But as the results became visible, business teams started bringing their own automation requests to the table. That shift — from IT-driven to business-driven demand — is a useful signal that the tooling has crossed the credibility threshold inside the organisation. W2M now has a dedicated process area for reviewing and redesigning workflows, which is the structural support that kind of cross-functional momentum actually needs to scale.
W2M’s invisible AI story is a practical counterargument to the idea that AI value lives mainly in customer-facing products. The 33,000 hours came from boring, high-volume back-office work — email triage, data enrichment, mobility quoting — handled by automation that employees barely notice is there. For builders evaluating where to deploy agentic workflows, that’s the signal worth paying attention to. Tools like n8n, Make.com, or LangChain-based pipelines can deliver this kind of integration without a full platform rebuild. For more on AI agents and automation tools, visit our AI Agents section.
Originally published at https://autonainews.com/world2meets-invisible-ai-saves-33000-hours-boosts-travel-efficiency/
Top comments (0)