SWARM Engineering has raised $10 million in an oversubscribed Series A round, as it looks to scale its AI-driven platform designed to transform how agri-food companies make operational decisions.
The San Francisco-based start-up, which positions itself as a decision intelligence company for agri-food and manufacturing, said the capital will be used to accelerate development of its operational AI platform, expand into new use cases, and scale go-to-market efforts. The funding will also support deeper integrations with ERP and supply chain systems, aimed at reducing time-to-value for customers.
The round was co-led by S2G Investments and AgRogue Growth Partners, with participation from Radicle Growth, Grit Road Partners, Middleland Capital, Open Prairie, Serra Ventures, and Trailhead Capital.
A response to “existential” pressure on operational decisions
SWARM is pitching its technology squarely at a sector under mounting strain. Agri-food businesses are contending with fragmented supply chains, volatile input costs, and rapidly shifting demand patterns – spotlighted most recently by a three-month war that has paralyzed the Strait of Hormuz.
“Margins that were once manageable are now existential,” the company said, arguing that disruptions – from trade route changes to labour shortages – are forcing operations teams to rethink how quickly they can adapt.
According to CEO Shail Khiyara, the question for many operators has shifted from whether to invest in AI to how quickly it can be deployed.
Yet much of the sector, he argues, is still relying on legacy planning tools built for a more predictable environment.
From hindsight to real-time decisions
SWARM’s answer is a platform built on what it calls an “operational ontology” of agri-food and manufacturing – embedding industry-specific decision logic, constraints, and variables directly into AI models.
Unlike more generic AI systems, Khiyara said SWARM’s platform is designed to reflect how decisions are actually made on the ground.
“In agri-food and manufacturing, every operational decision has a downstream consequence,” he said. “Most AI platforms learn your business over time. SWARM is different because it’s built on the ontology of these industries – the decision logic, the constraints, the relationships between variables that can take decades to accumulate.”
The system ingests structured and unstructured data from multiple sources – including ERP systems, spreadsheets, and real-time feeds – before running scenario analyses across thousands of variables in minutes.
The result, the company claims, is a shift from backward-looking planning to real-time, decision-ready intelligence.
Cutting planning cycles from days to minutes
SWARM points to early customer deployments as evidence of its impact.
At Ardent Mills, North America’s largest flour milling and ingredient company, rebuilding supply chain plans previously required a multi-day manual process across disconnected systems.
“With SWARM, their operations teams now simulate hundreds of logistics and supply chain scenarios in minutes,” Khiyara told AgNavigator. “Decisions that previously required days of manual rebuilding happen in a single planning session.”
Similarly, Springs Window Fashions, a global manufacturer with more than 9,000 employees, cut planning cycles by 40% and unlocked previously unseen working capital.
“SWARM didn’t just improve our planning process, it changed what’s possible,” said COO Oscar Bolaños. “We freed up working capital we didn’t know we had.”
Fast deployment reshapes the AI adoption equation
A key part of SWARM’s value proposition is speed of deployment.
Khiyara said customers typically reach full operational use within eight to ten weeks – far shorter than traditional ERP implementations, which can take 18 months or longer.
“That is not a pilot timeline. That is production deployment delivering real decisions within two months,” he said.
Crucially, the platform does not require deep ERP integration upfront. Instead, it is designed to work with fragmented and imperfect data environments – something Khiyara said reflects the reality of most agri-food operations.
“We meet customers where their data actually lives rather than requiring them to transform their infrastructure before we can begin,” he added.
Targeting high-complexity agri-food segments
Within agri-food, SWARM is initially focusing on protein production, grain processing, and food manufacturing – segments where operational complexity and decision frequency are highest.
“These are the environments where the cost of a slow or wrong decision is most immediate,” Khiyara said, citing factors such as perishability, biological variability, and multi-site logistics.
In addition to Ardent Mills the company has already deployed its platform with aquaculture firm Regal Springs.
Barriers to adoption: data and decision visibility
Despite growing investment in agri-food AI – AgFunder estimates global Agri-foodtech funding reached more than $16 billion in 2025, with around $5 billion flowing into deeptech – barriers remain.
Khiyara identified two persistent challenges: fragmented data and limited visibility into internal decision-making processes.
“Most organisations have more data than they realise, but it is fragmented and inconsistently labelled,” he said.
Equally, many lack a clear understanding of how decisions are actually made in practice, including informal constraints and bottlenecks. Mapping those processes, he said, often becomes one of the most valuable parts of a SWARM deployment.
From ‘nice-to-have’ to operational necessity
Investors backing the round argue the company is addressing a long-standing gap in agri-food technology.
“Agri-food has been underserved by technology for decades,” said Kirk Haney, managing partner at Radicle Growth. “SWARM is the first company we have seen that truly understands how operational decisions get made in this industry.”
Mike Wise, principal at S2G Investments, added that the platform’s domain specificity could prove decisive.
“Agri-food and manufacturing don’t have generic problems,” he said. “SWARM is building a platform with the domain depth and enterprise-grade foundation to become a defining player.”
For Khiyara, the ultimate test is how companies respond to disruption.
“When a logistics lane closes or a customer changes their order with 48 hours’ notice, the question is not whether you have data – the question is how fast you can turn that data into a decision-ready plan,” he said.
“SWARM is the difference between managing disruption and being managed by it.”
Strengthening agri-food ties
Alongside the funding, SWARM announced that Jason Trusley, SVP and chief strategy officer at Land O’Lakes, has joined its advisory board, bringing additional industry expertise.
As agri-food operators face an increasingly unpredictable operating environment, SWARM is betting that faster, higher-quality decision-making – powered by domain-trained AI – will shift from competitive advantage to operational necessity.
