Tackling Industrial AI’s Last Mile Problem: Our Investment in Arrakis
Summary
- Less than 10% of enterprises have moved AI agents into production at scale due to fragmented systems, legacy software, & unstructured data in physical operations.
- Arrakis addresses the last mile of AI deployment for industries like aerospace, defense, logistics, energy, & manufacturing using a full-stack approach with forward-deployed engineers & a model-agnostic sovereign platform.
- NYSE-listed customers in energy, logistics, & industrial sectors have achieved a 90% reduction in procurement cycle times & increased real-time visibility into manual processes.
- The company was founded by Rafael Quintanilla, formerly of Accel, alongside co-founders Haroun Beltaifa, Romain Fouilland, Michael Galkov, & Chester Thompson, who bring experience from firms including Palantir, Datadog, ASML, Uber, & KKR.
- Arrakis emerged from stealth with $38M in funding & partnered with Blossom Capital to support its growth.
- The firm deploys AI agents into existing systems without multiyear migrations while maintaining rigorous compliance, security, & data sovereignty where customer IP is never used for third-party training.
- Arrakis ties a significant portion of fees to customer outcomes, ensuring pricing aligns with value delivered, & has secured demand across eight countries within six months of launch.
Most AI investment so far has gone to workers who sit behind a desk, but those who run physical operations have seen little of it. This is not for lack of ambition - industrial companies want to deploy AI. However, they run on decades-old software, fragmented systems, and workflows held together by manual work and unstructured data. In that environment, most AI projects never reach production: less than 10% of enterprises have moved agents into production at scale.