New Way Now: LATAM Airlines clears the runway for AI innovation with Google Cloud

*Featured in this video:* Andrés Bucchi, Chief Data Officer at LATAM Airlines

*Executive summary:* LATAM Airlines, Latin America’s largest airline group, is elevating every single journey with data and AI, placing Google Cloud at the heart of its strategy. To meet increasing demand for personalized travel experiences, LATAM Airlines migrated its data platform to BigQuery and used Gemini Enterprise Agent Platform (formerly Vertex AI) to build Cosmos — a proprietary middleware platform that provides a standardized, secure, and cost-efficient process for deploying data and AI across its operations, without adding heavy suitcases of technical debt. Now, LATAM Airlines has the foundation it needs to fully embrace the latest AI technologies to fuel first-class personalization, paving the way for innovations like an AI concierge agent that helps passengers book and plan their trips using natural language interactions. Google’s Veo model is helping creative teams to generate highly personalized, localized video content for marketing campaigns, shrinking production timelines by up to 95% while cutting costs by 99%. BigQuery AI has also empowered employees with multimodal capabilities to tap into the full value of its data, transforming complex analysis that previously consumed millions of budget dollars every year into lean, convenient processes at a fraction of the cost.

*Challenge:* LATAM Airlines’ existing data foundation made it difficult to deploy AI without adding technical debt, limiting its ability to deliver personalized experiences to 100 million customers across more than 1,600 flights per day. Relying on slow, manual creative processes also prevented creative teams from moving fast enough to produce localized, timely, and hyper-personalized ad campaigns.

*Solution:* Partnering with Google Cloud, LATAM Airlines migrated its core data platform to BigQuery and then built Cosmos — a proprietary AI platform running as middleware over Gemini Enterprise Agent Platform. Cosmos serves as the mandatory, unified framework for deploying data and AI applications across every department, providing the security, compliance, and governance features and tools needed to automate, standardize, and manage AI projects and data pipelines at scale. This new foundation has also provided access to the best of Google AI, including Google’s state-of-the-art video generation model, Veo, and multimodal Gemini models, delivering powerful new capabilities that empower employees to explore data and supercharge creative processes faster than ever.

*Results:* This new foundation, powered by Google Cloud, is transforming the way LATAM Airlines provides first-class customer journeys. Cosmos is already charting new paths to value, driving new innovations like an AI concierge agent that helps passengers book and plan their trips using natural language interactions. Veo has supercharged creative processes, delivering highly personalized, localized video content up to 95% faster while cutting production costs by 99%. Finally, BigQuery AI has simplified complex analysis tasks — which previously cost millions per year — into convenient tasks anyone can perform for cents on the dollar.

*Key takeaways and highlights from our interview with Andrés Bucchi, Chief Data Officer at LATAM Airlines:*

→ “When you have these kind of creativity components, it's really hard to automate. What Veo allows us to do is set that as a technical piece. You just use your creativity and Veo just figures it out, and then you have your content. You can fine tune everything to the color of the seat, and the speed that you gain, it’s incomparable — and now you're talking about 90%, 95% reduction in lead times, in costs, 99%.”

→ “We have a proprietary platform called Cosmos. It's built as a middleware over [Gemini Enterprise Agent Platform]. You get security by design, you get all the observability, cost control, and with ADK, we can use the best model for the specific tasks. So, that has given us a flexibility and robustness to scale up AI, machine learning, and data tooling across every single piece of technology, and it opens the door to deliver with generative AI.”

→ “Having the data in your cloud environment and having it accessible through BigQuery is a huge win. You can use a single language and interface that has a single set of properties that are really flexible, and multimodal technology has allowed us to have structured ways of gathering KPIs and insights. Gemini allows us to look for this information in every single piece of unstructured data and we can create data products that then we can experiment with. If you want to analyze a very complex process with a lot of unstructured data that would set you back millions of dollars a year, now you can do that literally with cents on the dollar.”

*Google Cloud products used:* Gemini Enterprise Agent Platform (formerly Vertex AI), BigQuery, Veo, and Gemini

*Learn more:*
→https://www.latamairlines.com