By Kerry Wei, Denny Hua, and Erica Li
In 2025, use of AI agents was most commonly found in technology sectors. In comparison, operationally intensive industries like logistics, manufacturing, and others that underpin the physical economy have been slower to adopt this technology despite presenting some of its most compelling opportunities for value creation. Running these businesses requires complex, multi-step, multi-channel coordination and communication among customers, employees, suppliers, numerous internal systems, and more. Many also rely heavily on institutional knowledge that has historically remained undocumented.
Each of these individual interactions, coordination points, and knowledge hand-offs are often easy to overlook. At scale, though, they consume enormous amounts of employee capacity and are critical to keeping an organization running effectively. Reliable AI agents that seamlessly integrate into existing workflows are able to fundamentally change i) how work is executed and ii) the economic profile of a business.
Enter HappyRobot: AI Agents That Do Real Work
We had been tracking HappyRobot from the outside for more than a year and were introduced to Co-Founder & CEO Pablo Palafox and Head of Strategy & Ops Quili Peña Martinez-Avial by the Base10 team. What immediately stood out to us was their insistence on holding themselves to a higher standard than impressive demos: agents in production, doing measurable work and driving real business impact. That insistence perfectly captures their culture of “earning the right to do more work,” which resonates with some of the most operationally demanding enterprises inundated by AI applications.
HappyRobot started by building AI agents that handle high-volume operational tasks across channels, including voice, email, messaging, and digital workflows. Its natural, responsive voice capabilities allow agents to navigate complex conversations with a high-quality user experience. The agents plug directly into enterprise workflows and their supporting systems, so they can look up a shipment, update a record, schedule an appointment, and escalate to a human when judgment is required. Underneath sits Twin, HappyRobot’s operational data layer, capturing enterprise context and institutional knowledge from often-siloed systems of record (e.g., CRM, ERP). Twin codifies the knowledge found in every agent-to-human interaction, documented and undocumented process, workflow exceptions, and stakeholder relationship. This creates a reinforcing flywheel: with every interaction, agents better understand how the business operates, allowing them to work seamlessly and take on more work alongside employees without fundamentally changing how their customers operate.
Foundation models give agents the ability to reason, communicate, and act, but putting those capabilities reliably to work inside a large enterprise is a problem of orchestration, adaptation, and control. That is the layer that HappyRobot owns, and it’s where the company’s advantage compounds. In practice, it has enabled customers to put agents to work without reorganizing teams or retraining staff around a new system, the kind of friction that tends to stall enterprise AI adoption. HappyRobot agents meet the organization where it already operates, which is what makes them embedded operational infrastructure rather than another piece of software to manage.
HappyRobot has delivered measurable results at scale, improving key business metrics across large enterprises, including operational throughput, productivity, and cost efficiency. At WWEX, for example, six AI agents now track 25,000 loads per month from end to end, generating 10x the task throughput of a human team and an estimated six-figure annual ROI in operational savings.
That impact has driven rapid enterprise adoption. HappyRobot now works with 150+ enterprise customers, including DHL, Kuehne + Nagel, Naturgy, and Uber, and has grown 5x since raising its Series B late last year. Having first proven its platform in logistics, one of the economy’s most operationally demanding industries, HappyRobot continues to deepen those partnerships while expanding into insurance, energy and utilities, telecommunications, and other sectors where business-critical work still depends on complex coordination across fragmented systems. This expansion across industries and within existing customers is supported by two reinforcing advantages: its deployment capabilities strengthen with each implementation, while the agents operating become more effective as they accumulate deeper organizational context over time.
That momentum is why we’re proud to lead HappyRobot’s $150 million Series C, co-led by Eurazeo, bringing the company’s total funding to more than $200M.
Why This Fits Our Thesis
At Prysm Capital, we partner with founders building category-leading companies in large, growing markets. The AI agent market is expected to reach $235B+ by 2034, and we believe the winners will be defined not by model quality alone, but by depth of integration and ability to create measurable business value.
HappyRobot embodies our view. The company has established category leadership in a vertical where reliability is non-negotiable, by combining deep industry knowledge with the deployment capabilities required to operationalize agents at scale across critical workflows. With every implementation, HappyRobot can deploy faster, become more embedded, and deliver greater value over time as the company becomes a trusted partner.
This expertise is readily transferable across operationally intensive industries. As in freight, the same economic incentive is clear and creates urgency and durable demand: coordination is a permanent line item that AI can finally attack, and the companies that move first bank the savings as margin before competitors catch up.
HappyRobot is well-positioned at this inflection point, as the focus of enterprise AI shifts from theoretical intelligence to tangible execution and value creation. The company is laying the foundation for enterprise superintelligence, a shared operational understanding that deepens as agents work across the enterprise and capture context through every interaction.
Pablo Palafox, Javi Palafox, and Luis Paarup bring a unique, hands-on perspective to the team, combining deep research with real-world enterprise knowledge. Their approach has helped HappyRobot move beyond the role of a traditional vendor, truly embedding itself as a core extension of their customers’ operations.
We’re thrilled to support the team as they define what execution truly means in this new agentic era. This new funding accelerates investment in HappyRobot’s platform and team to support growing enterprise demand worldwide.
Disclaimer: For informational purposes only. Not an offer of any securities or of the investment advisory services of Prysm Capital, L.P. References in this piece are provided solely for illustrative purposes to highlight aspects of Prysm Capital’s investment approach and do not purport to represent all investments made by Prysm Capital. Past performance is not indicative of future results.