Summary

• Analysys Mason has acquired Stockholm-based Modulai for an undisclosed sum, adding specialist AI and machine learning engineering capability.
• Modulai develops production systems spanning generative AI, agents, computer vision, recommender systems, natural-language processing, and other applications.
• The deal reflects demand for advisers that can connect AI strategy and due diligence with engineering, deployment, and operation.

The boundary between technology consulting and software engineering is narrowing as businesses try to turn artificial intelligence strategies into operating systems, with Analysys Mason acquiring Swedish machine learning consultancy Modulai.

Financial terms were not disclosed, while the transaction adds a specialist engineering team to Analysys Mason and is intended to connect its strategy, transaction, and transformation work more directly with technical implementation.

Stockholm-based Modulai was founded in 2018 and develops tailored systems across generative AI, machine learning, computer vision, natural-language processing, recommender systems, robotics, and AI agents. Its work extends into production deployment rather than ending with strategy or experimentation.

That capability is increasingly relevant as organisations move beyond deciding whether they should use AI and encounter harder questions about architecture, integration, data, reliability, monitoring, security, and operational ownership.

Advice moves closer to the engineering layer

Analysys Mason has historically been associated particularly with telecommunications, media, technology, digital infrastructure, space, and investment advisory work, providing research, strategy, and transaction support to companies, investors, and public bodies.

Adding Modulai gives the consultancy a more direct route from identifying where AI could create value towards building the system that delivers it. Analysys Mason says the combined offering will include large-scale AI transformation, technical implementation, and AI due diligence across technology-intensive industries.

Technical due diligence is becoming more demanding as investors and acquiring companies try to understand whether an AI business owns defensible technology or has assembled its product largely from third-party models and infrastructure. Model quality, data rights, engineering maturity, compute dependencies, deployment processes, and operating reliability can all affect value without appearing clearly in revenue figures.

The same questions arise after an acquisition because an impressive demonstration may conceal weak monitoring, security, cost controls, or integration with customer systems. Assessing whether a product can run reliably in production requires engineering expertise alongside conventional market and commercial analysis.

Modulai’s current capabilities span model development, deployment, monitoring, retraining, model operations, hosting, and the software required around AI systems. That breadth gives Analysys Mason a way to carry advisory work further into implementation, where problems often emerge after a model has already been selected.

Implementation becomes the bottleneck

Generative AI has made model access relatively easy compared with the work required to deploy the technology inside an established organisation. Companies can buy services from multiple providers, but connecting them safely to internal data and workflows still depends on the condition of existing technology estates and the people capable of working across both software engineering and business processes.

Production systems also require more than the model itself because retrieval quality, permissions, evaluation, observability, tool access, latency, infrastructure cost, and failure handling can determine whether an application remains useful after the initial demonstration.

Large professional-services groups, systems integrators, specialist consultancies, and software companies are consequently competing for engineers who can turn AI projects into operating systems. Acquiring an existing specialist team can be faster than building that capability internally, particularly when experienced engineers remain scarce.

Modulai brings a track record that predates the current generative AI investment cycle, having operated since 2018 across sectors including financial services, life sciences, manufacturing, logistics, retail, energy, and real estate. Its own material describes more than 150 end-to-end AI projects across more than 100 clients and partnerships, although those figures remain company reported.

The acquisition also broadens Analysys Mason’s technical reach beyond the telecommunications and digital infrastructure markets with which it is most closely associated. Modulai’s work across manufacturing, finance, health, and other industries gives the combined group a wider base from which to sell implementation services.

Buying engineering capability does not remove the organisational problems that cause many AI projects to stall, because workflows, responsibilities, data quality, procurement, and employee adoption still have to change around the technology. The deal instead reflects how consulting demand is moving: strategy increasingly has to survive contact with the software, data, and infrastructure required to put the recommendation into production.