PwC and Multiverse Computing have expanded their alliance to promote Artificial Intelligence (AI) with a real impact on companies internationally. Following the strategic agreement signed three months ago, both companies have agreed that, in addition to working with Spanish companies, they will jointly launch projects in the United States, Canada, Germany, Brazil and Italy, taking advantage of the professional services firm's international network.
Multiverse Computing, a company based in San Sebastian and with offices in Paris, Munich, London, San Francisco, Toronto and Milan, is considered one of the leading companies in the compression of AI models and quantum software applied globally. Its CompactifAI technology allows large AI models to be compressed and optimized so that companies can apply AI models with lower investment and energy consumption.
The extension of the alliance was agreed at the meeting held by the top executives of both companies at the headquarters of Multiverse Computing in the city of San Sebastian, where they reflected on the impact that AI is having on companies, both nationally and in other geographies. César Calleja, partner in charge of Consulting at PwC Spain, and Enrique Lizaso, co-founder and CEO of Multiverse Computing, took stock of the first three months of the alliance, a period in which more than 30 working sessions have already been held with top-level executives in the Spanish market and in which the enormous interest of national companies in promoting solutions in which AI helps them has been detected to improve their income statement.
The meeting was also attended by Asier Atutxa, partner in charge of PwC's Public Sector and Northern Zone; Jon Toledano, partner responsible for audit in PwC's North Zone, and Armando Martínez, partner responsible for the Alliance with Multiverse Computing of the professional services firm. On behalf of Multiverse Computing, Rodrigo Hernández, Global Director of Generative AI, Gorka Otermin, Strategic Partnerships and Sales, and Ane Iturzaeta, Strategic Alliance Manager, participated.
Limited adoption of AI within organizations
One of the main reflections put on the table was the still limited adoption of AI at scale in the business world. Among the causes identified are the difficulty of accessing transactional systems, insufficient data quality, the absence of specific adoption programs, the lack of structural capacities or the lack of a clearly defined strategy, among others.
The 2026 Global CEO Survey, which PwC presented at the beginning of the year as part of the Davos Forum, reveals that a majority of senior executives (56%) admit that their companies were not yet obtaining significant returns on their investments in AI. The survey also found that despite the fact that a majority of companies are already experimenting with AI, only 12% of CEOs had achieved benefits in terms of both cost reduction and revenue growth. However, the same survey showed that the global CEOs who are reaping benefits, both in terms of cost reduction and revenue growth, are those who are integrating AI into their products and services, demand generation, and strategic decision-making.
AI with a focus on "monetization" of results
In this context, with a focus on the impact and "monetization" of results, the range of services developed by PwC and Multiverse Computing will allow companies in the different sectors of activity to deploy specific use cases with tangible effects on the business, while also guaranteeing parameters of sovereignty, security and efficiency. In fact, reflecting the market's interest in the synergies of collaboration between the two companies, both companies are working on more than a dozen opportunities on real and concrete use cases that will demonstrate the real impact of AI on the day-to-day life of organizations.
The solutions that are being developed make it possible to effectively address the different challenges that companies identify in their day-to-day work:
- Sovereignty of models, data and infrastructure.
- Information security in perfectly controlled environments.
- Improved accuracy of model outputs.
- Reduced latencies and model response/inference times.
- Deployment of the models in devices without coverage or with low performance.
- Significant cost reduction.