Arrakis Technologies: Bringing Agentic AI to Factories and Supply Chains

Founders and Team

Arrakis was founded by four people in January 2026 :

• Rafael Quintanilla — CEO and co-founder. Former Vice President at Accel, where he developed the firm's thesis on defense and industrial resilience. He left Accel to start Arrakis .
• Haroun Beltaifa — Co-founder, formerly of Palantir .
• Romain Fouilland — Co-founder, formerly of Palantir .
• Mikhail Galkov — Co-founder, formerly of Delivery Hero .

The broader founding team also includes alumni from Revolut, Datadog, and ASML .

Investors and Backers

Arrakis raised a total of approximately $38 million across two rounds :

The company's post-money valuation stands at $140 million .

Notable angel investors include:

• Olivier Pomel — CEO and founder of Datadog
• Olivier Godement — OpenAI's Head of Business Products
• Junaid Hussein — Founder of Cambridge Aerospace

How the Platform Works

Arrakis combines three layers to deliver AI into industrial environments :

1. Forward-deployed AI engineers who embed directly at customer sites to design and build custom agents.
2. A model-agnostic AI platform that lets customers use models from OpenAI, Anthropic, Mistral, or open-source alternatives, switching between them per task .
3. Applied AI research tuned to industrial operations, focused on handling unstructured data (PDFs, emails, spreadsheets) and integrating with existing enterprise resource planning systems (SAP, Oracle, Infor, IBM, Workday).

The company claims its model-agnostic approach delivers a 2–4x improvement in output quality while cutting token costs by roughly 70% .

Key platform capabilities include:

• Data intake and integration with existing enterprise systems (no rip-and-replace required)
• Workflow logic and governance for auditable, traceable actions
• An execution hub and strategic command center for operations teams
• SOC 2, ISO 27001, and GDPR compliance; deployable on customer cloud or on-premise

Early Enterprise Results

Arrakis says it already has five enterprise customers, including NYSE-listed companies .

Specific results reported:

• 90% reduction in procurement cycle times for one client .
• Cash-flow visibility improved from monthly to daily at a NYSE-listed shipping company, achieved by rebuilding a spreadsheet with AI populating data and learning from operator corrections .

The startup ties roughly 50% of its fees to hitting performance targets, an outcome-based pricing model that is unusual in the industry .

Expansion Plans

With the new capital, Arrakis plans to :

• Triple its headcount from roughly 15 employees.
• Open offices in New York and the Middle East, focusing on markets where industrial transformation is a government priority.
• Continue targeting family-controlled businesses, which CEO Quintanilla says make faster purchasing decisions and think longer-term .

Competitive Positioning

Arrakis positions itself as a faster, more flexible alternative to Palantir — which Quintanilla calls "a 20-year-old company with technology starting to become legacy" — and traditional consulting firms such as Accenture and BCG, which charge for hours rather than outcomes . It also competes with other industrial AI firms like PhysicsX ($300M raised) and Jeff Bezos-backed Prometheus, but Quintanilla argues that Arrakis focuses on operational workflows surrounding core engineering rather than engineering simulation or product design .