BEIJING, May 8, 2026 — Following its RMB 1 billion strategic round in March, robotics company ROBOTERA has raised over USD 200 million in a new financing round led by SF Group. The round also saw participation from leading financial investors including HSG, IDG Capital, Hillhouse Investment, CICC Capital, Jingming Capital, SparkEdge Capital, Luxin Venture Capital Group, Unite Pioneers Capital, and Longqi Investment; alongside major industrial partners such as KENGIC, Dongfeng Asset Investment, ICBC Capital, and funds affiliated with China Unicom. Existing investors Tsinghua Holding Tiancheng Asset Management and Horizon Investment continued to increase their stakes. Investor demand significantly exceeded the initial fundraising target.
ROBOTERA M7 humanoid robot performing small parcel induction and sorting in a logistics environment
ROBOTERA now brings together top-tier financial institutions including CDH Venture and Growth Capital, HSG, IDG Capital, Hillhouse Investment, Gaocheng Capital and CICC Capital, alongside a broad base of industrial investors such as SF Group, Alibaba, Geely Capital, BAIC, Dongfeng Asset Investment, Lenovo, Haier, Golden Resources Group, Singtel Innov8, China Unicom-affiliated funds, Woori Venture Partners, a venture capital subsidiary of Woori Financial Holdings, ICBC Capital, KENGIC and a leading South Korean technology company. These partners contribute real-world deployment scenarios and commercial demand, supporting large-scale adoption.
This strong backing reflects confidence in ROBOTERA’s technology path and its ability to deliver productivity at scale. The company has achieved the first product-market fit (PMF) in the embodied intelligence sector, with deployments across more than ten logistics centers in collaboration with China Post and SF Group. In Q2 2026, ROBOTERA initiated thousand-unit deliveries, with growth exceeding 300%.
Beyond deployment, ROBOTERA has built a fully in-house robotics hardware system, with over 95% of core components developed internally, spanning actuation systems and humanoid platforms, forming a robust foundation for real-world mobility and manipulation tasks.
A key focus is its pioneering full direct-drive dexterous hand architecture, the first of its kind in the industry, enabling high-precision, adaptable, and durable manipulation in logistics and industrial environments. Its reliability has been validated through long-term real-world deployment.
ROBOTERA’s hardware system has been adopted by leading global technology companies and research institutions, including Boston Dynamics, NVIDIA, and Apple. The system works in close coordination with software in real-world environments, with deployment feedback enabling continuous performance optimization.
With proven deployment in logistics and ongoing expansion into automotive, electronics, and service industries, ROBOTERA is entering a phase of rapid commercialization. The company will continue scaling real-world robotic applications across global markets.
NEW YORK, May 7, 2026 — Balcony, an innovative company building modern data infrastructure for the U.S. property market, today announced it has raised a $12.7 million seed round led by Blockchange Ventures, bringing its total raise to $14 million. The financing will be used to accelerate the development of “digital rails” for America’s real estate economy. The Keystone infrastructure is currently helping government agencies in the U.S. manage and secure over $400 billion in property value on its platform.
The funding will scale Balcony’s engineering and go-to-market teams and expand the deployment of its platform across county and state governments nationwide.
Building the Foundational Infrastructure for Real Estate More than 3,000 county offices maintain the land records that form the legal bedrock for trillions of dollars in U.S. real estate. For decades, this information has been fragmented across local systems, creating friction and risk for the entire economy. For the first time, Balcony’s platform integrates with county systems to transform historical records into the digital rails for the property market, creating a structured and connected data layer that can power a modern, secure economy.
This infrastructure provides a 360-degree parcel view, empowering key partners such as title insurers, mortgage lenders, and capital markets to operate with greater speed and certainty. The company’s mTrace platform builds on these digital rails to provide intelligence-driven threat detection for government agencies.
Proven at Scale in Partnership with Government Balcony recently signed a five-year contract with the Bergen County Clerk’s Office in New Jersey to digitize and bring 370,000 property parcels onto its platform, representing approximately $240 billion in real estate value.
“For counties like ours, modernizing how land records are organized and accessed is critical,” said John Hogan, County Clerk of Bergen County, New Jersey. “Balcony’s platform works alongside the systems we already use to help us organize decades of records in a way that improves transparency and makes information easier for both our office and the public to access.”
“We have a profound respect for the public records that counties steward,” said Gregg Lester, co-CEO and President of Balcony. “Our role is to partner with these public servants to build a modern, secure, and connected data layer upon that bedrock. We are not replacing their critical systems, but rather building the digital rails alongside them to ensure these foundational records can power a more transparent and secure market for the next century.”
The Investor Vision: A National Security Imperative for Digital Rails For lead investor Blockchange Ventures, the financing follows its successful playbook of rebuilding foundational infrastructure, a strategy proven with its early backing of fintech leader Figure Technologies. The firm believes that the decision to back a nationwide upgrade of property records, a historically fragmented and monumental undertaking, is only viable now because a powerful new catalyst has emerged.
“Property ownership is a pillar of our economy, and in today’s world, it’s also a matter of national security,” said Ken Seiff, Managing Partner at Blockchange Ventures. “The drive to build these digital rails is imperative because our fragmented, century-old system is vulnerable. Balcony’s unique ability to create a verifiable and connected view of land records is the gamechanger, empowering governments to protect against fraud and monitor foreign ownership, which in turn builds the foundation of trust required to finally modernize the rails for the entire real estate economy.”
About Balcony Balcony builds the digital rails for America’s property records. By partnering with government agencies, the company is creating the foundational infrastructure to secure and modernize the real estate economy. This trusted data layer empowers partners, including title insurers, mortgage lenders, and financial institutions, to operate with greater confidence and security. Balcony is headquartered in the Greater New York area. To learn more about Balcony visit https://balcony.technology/.
DENVER, May 7, 2026 — Industrious Ventures congratulates Lunar Outpost, the leader in off-planet mobility, in-space infrastructure, and space-based autonomous platforms, on today’s announcement of its oversubscribed $30M Series B. Industrious Ventures led the round with participation from Type One Ventures, Eniac Ventures, Reliable Equity, and others.
Justin Cyrus (CEO), AJ Gemer (CTO), Forrest Meyen (CSO), and Julian Cyrus (COO) founded the company to develop next-generation mobility and autonomous systems that enable reliable, scalable operations on the lunar surface and in orbit. The funding accelerates production and development of its rover systems, expanded manufacturing capacity, and deployment of mission-ready systems for NASA Artemis and U.S. national security programs.
As global efforts accelerate toward a permanent human presence in space, mobility and autonomy are foundational requirements to support sustained operations in extreme environments and the broader space economy. Lunar Outpost is a key provider in this category, with several different classes of scalable rover systems validated through flight heritage and representative environmental testing.
Lunar Outpost is also advancing autonomous robotics and coordination software through its STRATFI-supported MARS platform, enabling resilient multi-vehicle operations in GPS-denied environments. The capabilities are being tested across lunar, orbital, and terrestrial applications, expanding the company’s reach into broader defense and space-based autonomy systems.
“Lunar infrastructure development will come in phases over the next 10 years, shifting from site preparation to development to sustainment,” said Taylor Sargent, Partner at Industrious Ventures. “That progression depends not just on mobility, but on systems that can operate autonomously in extreme environments. Lunar Outpost has demonstrated the ability to execute in the field while building capabilities that translate directly to how work gets done on the lunar surface, in cis-lunar space, and beyond.”
“As activity on the lunar surface and space accelerates, the systems that can operate reliably and scale will define what’s possible,” said Mason Angel, Founding Partner at Industrious Ventures. “Mobility is a critical foundation for everything that follows and increasingly requires autonomy to operate in complex environments. Lunar Outpost is building with that full system in mind.”
Lunar Outpost has developed and deployed multiple generations of its Mobile Autonomous Prospecting Platform (MAPP) rover, supporting applications including resource prospecting, infrastructure development, and sustained surface operations. Building on that foundation, the company is advancing multiple next-generation lunar terrain vehicles as part of NASA’s Lunar Terrain Vehicle Services (LTVS) program for Artemis, reflecting a broader shift toward commercially developed, rapidly deployable mobility systems.
“From the beginning, our focus has been on building systems that can operate reliably in the harshest environments and scale with demand,” said Justin Cyrus, Founder and CEO of Lunar Outpost. “This next phase of growth allows us to accelerate deployment of the robotic and mobility infrastructure required for sustained operations beyond Earth and the broader expansion of the space economy.”
Industrious Ventures backs founders building the core technologies required for the next era of industrial capability on and beyond Earth. Lunar mobility represents a critical unlock for that future, enabling infrastructure deployment, resource utilization, and human exploration. With increasing demand from government and commercial stakeholders, Lunar Outpost is positioned to define how work gets done both on the lunar surface and in space in the years ahead.
About Lunar Outpost
Lunar Outpost is the leader in space mobility and infrastructure, developing advanced robotic systems for extreme environments. From enabling the first commercial rover on the Moon and making oxygen on Mars to driving NASA’s Lunar Terrain Vehicle program, Lunar Outpost is accelerating the new space economy for the betterment of life here on Earth. Leading the industry with 8 missions launching to the Moon before 2030, Lunar Outpost is scaling the core technologies that power the industrialization of space. For more information, visit lunaroutpost.com.
About Industrious Ventures
Industrious Ventures backs founders reinventing the world’s most critical industries—from the deep sea to deep space. With evergreen capital and long-term conviction, the firm partners with deep-tech builders from early development through scale-up and production. Industrious invests across aerospace, national security, energy, compute, manufacturing, and other foundational sectors where innovation strengthens the industrial economy. Learn more at www.industrious.vc.
The viral clean-label french fry brand grew over 1,100% in 2025 and is on track to nearly quadruple in size again in 2026 as it launches nationwide at Target, expands at Whole Foods Market, and opens new retail doors.
WASHINGTON, May 7, 2026 — Jesse & Ben’s, the seed-oil-free french fry brand redefining the frozen potato category, announced the close of an oversubscribed Series A financing round led by Greycroft. The capital will fuel the brand’s continued retail expansion, supply chain investment, and leadership team build-out as the company scales nationally.
The round was led by Greycroft, with investment from the firm’s consumer brands fund, championed by Partners Brian Bustamante-Nicholson and Eric Ryan. Ryan, co-founder of Method and Olly, will join the company’s board. Jesse & Ben’s reflects Greycroft’s focus on backing founder-led consumer brands with cultural momentum that are reinventing legacy categories.
“Jesse and Ben are building exactly the kind of consumer brand we look to support, one that brings new energy to a large, established category,” said Brian Bustamante-Nicholson, Partner at Greycroft. “We believe their growth reflects strong consumer demand for simpler, better ingredients, and we’re excited to support the team as they continue to scale.”
Alongside lead investor Greycroft, the round saw new participation from Rich Products Ventures, the corporate venture arm of the global, family-owned food company, Rich Products, along with existing backers Willow Growth Partners, Sling Ventures, Midnight Venture Partners, and grt sht ventures.
Additional investors in the round include leading consumer operators and founders, including Andrew Abraham (Orgain / Humble Growth), Carter Comstock (Freshly / Comstock Growth), Allison and Stephen Ellsworth (Poppi), Nic Jammet (Sweetgreen), and Jordan Brown (Hu Chocolate), as well as creator-investors including Lauryn and Michael Bosstick (Dear Media / The Skinny Confidential) and many more.
“Jesse and Ben’s is doing exactly what we look for in the future of food: taking a massive, legacy category like frozen fries and completely reinventing it with a clean-label, brand-led approach. Their incredible performance and the way they’ve captured the cultural moment around better-for-you ingredients in under two years is incredible. We are thrilled to support the team as they scale this disruption nationwide,” said Brian Bernstein, Principal at Rich Products Ventures.
Jesse & Ben’s was founded by Jesse Konig and Ben Johnson, two restaurant operators turned CPG founders who launched their namesake brand on retail shelves in June 2024. The brand has scaled at a pace rarely seen in frozen food. Jesse & Ben’s grew over 1,100% in 2025 and is on track to grow another 300-400% in 2026.
The brand was put on the map through early nationwide launches at Sprouts Farmers Market and Whole Foods Market. Momentum continues to accelerate following a nationwide rollout at Target, regional rotations at Costco, and additional retail launches underway. Jesse & Ben’s has become the #1 brand in the frozen potato category at multiple retailers, surpassing legacy competitors and private label options.
A majority of the funding from the round will go toward scaling supply chain capacity to keep pace with national demand, supporting continued retail expansion across conventional grocery, club, and mass channels, and building out the team. Jesse & Ben’s recently brought on Audrey Burger (Primal Kitchen / Kraft Heinz) as Founding President, alongside other key hires from notable brands such as Siete, to help guide the brand in this new phase of growth.
Jesse & Ben’s launched with three restaurant-quality varieties made from just russet potatoes, 100% grass-fed beef tallow or avocado oil, and seasonings, bringing an elevated french fry experience into homes nationwide. The company is expanding the lineup in 2026 with new launches, including:
Sweet Potato Tallow Fries (just sweet potatoes, 100% grass-fed beef tallow, and sea salt) expands from an exclusive test-launch at Sprouts to multiple national retailers.
Crinkle Cut Beef Tallow Fries will be joining the core lineup this summer at select retailers.
Both products reflect Jesse & Ben’s commitment to giving people delicious versions of their favorite all-American foods with minimally-processed ingredients:
“When we started Jesse & Ben’s, the goal was simple: turn junk food into joy food. Two years in, we’re watching real people pick our bag over the brands they grew up with, and we’re outselling legacy brands on shelves we never imagined we’d be on. This round gives us the firepower to keep launching products people actually want, to scale our supply chain, and to bring in the best operators in CPG. We couldn’t have asked for a better group of partners to do it with,” said Jesse Konig, Co-Founder & CEO, Jesse & Ben’s.
About Jesse & Ben’s Jesse & Ben’s is a premium frozen french fry brand founded in 2024 by co-founders Jesse Konig and Ben Johnson. After a decade running Mid-Atlantic restaurant brand Swizzler, the two set out to fix what they saw as the most overlooked category in the grocery store: frozen fries. Every Jesse & Ben’s product is made with simple, real ingredients, zero seed oils, and the kind of craft you’d expect from a chef-run kitchen, not a frozen aisle. Today, Jesse & Ben’s products are available in more than 3,500 stores nationwide, including Whole Foods Market, Sprouts Farmers Market, Target, Thrive Market, Fresh Thyme, MOM’s Organic Market, and hundreds of independent retailers. Real Fries, Real Guys. Learn more at jesseandbens.com.
Institutions can finally license rare, high-fidelity datasets directly to AI builders, with no custodial broker in the data path
ST. PAUL, Minn. and HOUSTON, May 7, 2026 — KnectIQ Inc., pioneer of the programmable sovereign trust fabric, and VeridatAI Inc., the marketplace for compliant data exchange, announced a Strategic Partnership to deliver Custody-Free Data Exchange Infrastructure: a new category in which institutional datasets move directly from owner to buyer, never held by a custodial intermediary. VeridatAI will integrate KnectIQ’s patented SelectiveTRUST® as the trust fabric of its custody-free marketplace. Neither company sits in the data path.
The broken convention
Every prior marketplace has required data to enter the operator’s environment to be licensable. That assumption produces outcomes institutions cannot accept: IP exposure, GDPR and CCPA risk, copies in third-party clouds, provenance claims that won’t survive legal challenge. Clinical, genomic, financial, and research datasets remain unlicensed while AI developers scrape lower-fidelity alternatives. Incumbents cannot solve this; their business models require custody to function.
The new category
VeridatAI orchestrates discovery, licensing, agreements, and cryptographic provenance watermarking. SelectiveTRUST establishes a device-rooted, single-use trust binding for each authorized exchange. Data moves directly from owner to buyer, governed but never touched by the trust fabric above. Trust is ephemeral and provable, not persistent and assumed.
What this unlocks
The datasets that will define the next decade of AI sit inside hospitals, banks, research universities, and industrial systems. Under custodial architectures, owners will not release them, forcing AI developers onto scraped or synthetic substitutes. Custody-Free Data Exchange Infrastructure changes the economics: owners license data directly to AI developers, with provenance that survives downstream use.
The architecture works because of a specific dependency. Anthropic’s Mythos System Card documents AI agents that autonomously discover zero-day vulnerabilities and chain exploits to escape sandboxes, defeating conventional cryptographic and identity stacks. Any marketplace built on a conventional trust fabric inherits that exposure. SelectiveTRUST does not: KnectIQ’s globally patented methods architecturally prevent agents from forging a binding. The unlock and the immunity are the same architectural decision.
Commercial availability
The platform is available to institutional partners through VeridatAI’s beta, with general availability in Q3 2026. Initial deployments span healthcare, financial services, and advanced manufacturing.
“The world’s most valuable data has been held back by the assumption that licensing requires custody,” said Ken Morris, Founder and CEO of KnectIQ. “SelectiveTRUST and VeridatAI break that assumption. Without a trust binding, no communication channel exists: no binding, no communication, no access, no execution. The data AI’s next decade needs can finally move on terms owners accept.”
“The datasets that will define AI’s next decade sit inside universities, hospitals, banks, and factories,” said Michael McLaughlin, CEO and Co-founder of VeridatAI. “Custody-Free Data Exchange Infrastructure removes the broker from the data path. Data moves directly from owner to buyer, with every exchange cryptographically provable and provenance-watermarked. We give institutions a way to turn proprietary data into a liquid asset without turning it into a liability.”
About KnectIQ
KnectIQ Inc. pioneers the programmable sovereign trust fabric category. Its patented SelectiveTRUST® technology establishes device-rooted, ephemeral, quantum-resistant trust relationships that protect data, networks, and AI systems across domains and mission environments. Headquartered in St. Paul, Minnesota. www.knectiq.com.
About VeridatAI
VeridatAI Inc. operates the global marketplace for compliant data exchange, powering the AI economy with verified, high-fidelity datasets and a zero-custody Trust Infrastructure. Headquartered in Houston, Texas. www.veridatai.com.
24/7 indoor golf simulator franchise with 65+ territories nationwide will deploy new capital toward headquarters expansion, franchisee- and consumer-facing technology, and continued national growth.
COVINGTON, Ky., May 7, 2026 — Another Nine, the 24/7 indoor golf simulator franchise redefining how people access and experience premium golf, today announced the closing of a $2 million investment round funded entirely by the company’s existing investor base.
The capital will be used to expand Another Nine’s Northern Kentucky-based headquarters team, accelerate development of the company’s franchisee- and consumer-facing technology — including its proprietary A9OS software — and support the continued buildout of its national franchise network.
Simulator Suite at Another Nine Indoor Golf
Exterior of Another Nine Montgomery in Cincinnati
Another Nine has signed more than 65 franchise territories across 16 states since awarding its first territory in July 2025. The brand operates two company-owned locations in Cincinnati, with the first franchisee-owned locations slated to open later this spring. The company reported operational profitability in the first quarter of 2026.
“Beyond capital, our existing investors bring decades of franchise experience and chose to keep building with us — we appreciate their conviction and alignment with our vision,” said Ethan Grob, Co-Founder and CEO of Another Nine. “With this capital, our focus remains the same: building a brand that golfers of all skill levels love, and supporting our franchisees in delivering that experience everywhere.”
Another Nine’s franchise model is built around all-private simulator suites, 24/7 autonomous access requiring no onsite staff, and Trackman gameplay in every suite. Guests reserve by the hour without a required membership, with optional membership tiers, leagues, lessons, and group events available. The proprietary A9OS platform handles bookings, memberships, access and facility management, and reporting.
Founded in 2023 and headquartered in Covington, KY, Another Nine opened its first company-owned location in the Columbia-Tusculum neighborhood of Cincinnati in March 2024 and its second in Montgomery in February 2026. The company is not disclosing investor names or valuation terms.
About Another Nine
Another Nine is a 24/7 indoor golf simulator franchise born in Cincinnati, headquartered in Covington, KY. With all-private simulator suites, Trackman gameplay, and proprietary franchise technology, Another Nine is redefining how players experience premium golf — on their terms. The brand has signed more than 65 franchise territories across 16 states. Learn more at anothernine.com.
GALWAY, Ireland, May 7, 2026 — FotoNation® LTD., a Galway-based company, announced it has closed its Pre-A round led by Enterprise-Ireland and Silicon Gardens, with participation from other notable Angels.
This capital will accelerate FotoNation’s mission to develop its TriSilica® product: an ultra-low-power, perception-AI chip. Specifically, the funds will be used to progress the completion of the MPW prototype, TS-210.
” This investment, from a leading European VC as well as from the Irish Government, marks a pivotal moment for our company. On behalf of the management team, I thank our investors for their shared conviction in our long-term potential, cementing FotoNation as a leader in the European AI hardware race”, said Petronel Bigioi, CEO, FotoNation.
Since its Q3/2024 reincarnation, FotoNation, with many of its original team members, engaged in numerous design wins and reclaimed its market position as a leader in optimized ultra-low power perceptual-AI solutions. Investor support validates market demand for a dedicated hardware sensor-processing chipset.
“FotoNation is building ultra-efficient, hardware-accelerated edge AI that fuses vision and multimodal sensing on device,” said Gregor Rebolj, Managing Partner, Silicon Gardens. “Their technical depth and focus on privacy-aware experiences position them to help define the post-smartphone era of AI companions, from Physical AI to perceptual intelligence.”
Enterprise Ireland, the Irish government’s enterprise development agency, co-led the round, further cementing FotoNation’s commitment to the West of Ireland.
Leo McAdams, Head of Advanced Manufacturing and Lifesciences at Enterprise Ireland said, “Enterprise Ireland is proud to support FotoNation’s next phase of growth, reinforcing Ireland’s advanced technology and semiconductor ecosystem. This investment underscores our commitment to helping ambitious Irish companies scale globally, create high–value jobs across all regions, and build long–term economic resilience. We look forward to working closely with FotoNation as it seeks to expand internationally.”
About FotoNation: FotoNation Ltd. is an Irish AI company headquartered in Galway, with a development center in Brașov, Romania. FotoNation designs silicon and IP built around a single conviction: eliminate unnecessary data before inference begins. Its TriSense™ IP core brings together neural ISP, sensor fusion, and AI for perceptual cognition under extreme power constraints. Its TriSilica® chip family adds industry-first high-capacity bonded memory, translating that architecture into production silicon for devices that perceive, adapt, and act. FotoNation’s technology is embedded in more than 4 billion products worldwide, contributing to dozens of top-tier tape-outs across the consumer and edge AI markets. For more information, please visit: www.fotonation.com.
About Enterprise Ireland: Enterprise Ireland is the Irish Government agency that works with Irish enterprises to help them start, compete, scale and connect and win export sales in global markets. Enterprise Ireland partners with entrepreneurs, Irish businesses, and the research and investment communities to develop Ireland’s international trade, innovation, and competitiveness. In this way, we support sustainable economic growth and regional development and help create and sustain employment in Ireland. www.enterprise-ireland.com
About Silicon Gardens: Silicon Gardens is a venture capital firm, run by entrepreneurs who have built and scaled their own companies. They are dedicated to backing the next generation of high-growth innovative technology businesses. Silicon Gardens partner closely with founders to help them build durable, category-defining companies. https://www.silicongardens.com/
SAN FRANCISCO, May 6, 2026 — Corgi announced today it has raised $160 million in Series B funding at a $1.3 billion valuation, advancing its mission to build the first AI-native, full-stack insurance platform for startups. The company was founded by Emily Yuan and Nico Laqua.
Corgi Series B Raise
The round was led by TCV, with participation from both existing and new investors including Oliver Jung, Leblon Capital, Kindred Ventures, Repeat VC, Zone 2 Ventures, Audeo Ventures, Quadri Ventures, First Order Fund, Vocal Ventures, Maiora Ventures, Nordstar, Seven Stars Ventures, Hexa Capital, Alpha Square Group, GSBackers, OurCrowd, Alumni Ventures, Global Growth Fund, and other strategic investors.
This financing builds on Corgi’s previously announced $108 million across its seed and Series A rounds, backed by investors including Y Combinator, Kindred Ventures, Contrary, and more. Since receiving regulatory approval in July 2025, Corgi has launched what it describes as the first AI-native, full-stack insurance platform purpose-built for startups. This new funding brings Corgi’s total amount raised to over $268 million.
The new capital will support the continued expansion of Corgi’s startup insurance products, including broader coverage, deeper distribution, and ongoing investment in the AI systems that power underwriting, claims, and policy operations. The company is also moving into new verticals, beginning with trucking, where it plans to bring faster quoting, more adaptive risk models, and coverage aligned with real-world operations.
“Insurance is one of the largest industries in the world, but it’s still built on infrastructure from centuries ago,” said Emily Yuan, co-founder and COO of Corgi. “We started with property management and are expanding into trucking insurance, payroll, and small business, automating some of the hardest workflows in the real economy.”
Corgi’s long-term vision is to modernize one of the most complex sectors in financial services. Traditional insurance is fragmented across TPAs, MGAs, reinsurers, and carriers, resulting in slow processes, disconnected systems, and delayed decision-making. Corgi is rebuilding this stack from the ground up to enable faster decisions, streamlined operations, and real-time coverage.
“Where other companies might take the boring but safe path, Corgi will always dream bigger, accomplish more, and take more swings for the fences,” said Nico Laqua, co-founder and CEO of Corgi. “We will for sure always be the most passionate, genuine, curious, and ambitious of any company.”
About Corgi Corgi is an AI-native, full-stack insurance carrier built for startups. As a licensed carrier, Corgi designs and manages insurance end-to-end, using modern infrastructure and AI to power underwriting, policy management, and claims. The company delivers fast, flexible coverage tailored to how startups operate and scale.
– JuliaHub recauda 65 millones de dólares en su ronda de financiación Serie B y lanza Dyad 3.0, que lleva la IA agencial a los gemelos digitales industriales
Dyad, la primera plataforma de IA agencial del mundo para la ingeniería de hardware, lleva la IA física al diseño y las pruebas de sistemas complejos, reduciendo el tiempo de I+D de meses a tan solo unos días.
CAMBRIDGE, Mass., 6 de mayo de 2026 — Hoy, JuliaHub anuncia el lanzamiento de Dyad 3.0 y una ronda de financiación Serie B de 65 millones de dólares liderada por Dorilton Capital, con la participación de General Catalyst, AE Ventures y el inversor tecnológico y exconsejero delegado de Snowflake, Bob Muglia. Dyad supone un cambio fundamental en la forma en que se diseñan y construyen los sistemas físicos, al incorporar agentes de IA autónomos al diseño y las pruebas digitales de maquinaria industrial. Desde bombas de calor hasta satélites y semiconductores, los equipos de ingeniería pueden reducir los ciclos de diseño, pruebas y construcción de meses a minutos. Varias empresas de la lista Fortune 100 ya están utilizando Dyad y Julia en diversos sectores industriales como el aeroespacial, el gubernamental, el automotriz, el de climatización y el de servicios públicos.
Agent-created, live running dashboard of a model: Agent-created interactive, live simulation dashboard of a satellite’s photovoltaic power system along with orbit simulation. Model and dashboard both created with the Dyad platform, without needing to switch tools.
Cooling circuit in a typical data center: Cooling circuit models are used to size the chiller, study its performance under typical load or to tune the control system. The agent is able to construct and run various load profiles, to assess the controller’s performance.
Daniel Freeman, quien lideró la ronda de financiación Serie B para Dorilton Capital, comentó: “El modelado de sistemas es una de las capas estratégicamente más importantes de la arquitectura de ingeniería nativa de IA, ya que es donde convergen la física, la lógica de control y la IA. JuliaHub ha creado algo extraordinario con Dyad: una plataforma que no solo modela sistemas, sino que los compila, guiando a los ingenieros desde el concepto hasta el código de control de producción en un único entorno. Creemos que JuliaHub tiene el potencial de convertirse en una de las empresas líderes en IA física, y nos enorgullece apoyar al equipo en su rápida expansión hacia el mercado de Dyad”.
El “problema difícil” de la innovación en hardware
La ingeniería física representa uno de los sectores más grandes que aún no se ha beneficiado plenamente de la revolución de la IA. Si bien herramientas como Claude Code, Codex y Gemini han transformado el desarrollo de software, los ingenieros industriales siguen limitados por herramientas obsoletas. McKinsey estima que se necesitará una inversión acumulada de 106 billones de dólares hasta 2040 para satisfacer la necesidad de infraestructura nueva y actualizada. Los ingenieros que planifican y construyen estas actualizaciones necesitan una solución que les permita avanzar al ritmo del software mejorado con IA. Ahí es donde entra Dyad.
Dyad ofrece a los equipos de ingeniería un entorno basado en IA para modelar, probar y validar sistemas industriales: un ejemplo de Claude Code aplicado al mundo físico. Dyad 3.0 se lanza hoy y se basa en Dyad 1.0, lanzado en junio de 2025, y Dyad 2.0, lanzado en diciembre de 2025. Dyad conecta agentes autónomos con simulaciones físicas escalables, controles rigurosos, análisis de seguridad y la capacidad de generar código para sistemas embebidos, uniendo así el software con el mundo real. Ya sea una planta de tratamiento de aguas residuales o un automóvil, ya no se requiere un doctorado para desarrollar gemelos digitales altamente detallados, ajustar controladores para escenarios de implementación especializados e iterar en diseños de hardware para construir la máquina más eficiente desde el principio.
“No se trata de ayudar a los ingenieros a completar una pequeña tarea a la vez. Se trata de ingeniería automatizada a gran escala, donde los equipos pueden proporcionar una especificación completa a Dyad y este diseña el sistema completo. Especificación de entrada. Diseño de salida”, afirmó Viral Shah, consejero delegado de JuliaHub.
Gemelos digitales con aprendizaje automático científico
Los agentes basados en la nube de Dyad están diseñados para escanear continuamente el conocimiento científico mundial y así mejorar constantemente los modelos. Las pruebas de laboratorio automatizadas con IA se están ampliando para garantizar que los modelos se ajusten a la realidad física. La transmisión de datos combinada con el Aprendizaje Automático Científico (SciML) permite que los modelos evolucionen automáticamente a medida que el sistema aprende del mundo real. El ecosistema y el lenguaje de simulación de Dyad ofrecen una base sobre la cual todos estos aprendizajes se transmiten a los ingenieros para verificar los procesos, determinar si las suposiciones coinciden con los requisitos del cliente y actuar como intermediarios que garantizan la seguridad del producto final. El diseño de Dyad implica que los ingenieros no tienen que escribir cada línea de código para probar millones de diseños, a la vez que les proporciona las herramientas adecuadas para asegurar que los aviones permanezcan en el aire.
Prith Banerjee, vicepresidente sénior de Innovación de Synopsys, comentó sobre la colaboración con JuliaHub: “Dyad está transformando la ingeniería de sistemas al combinar IA científica, modelado automatizado y una potente canalización de compilación en un flujo de trabajo unificado. Integrada con el software de simulación Ansys TwinAI™ de Synopsys, permite crear gemelos digitales híbridos de alta fidelidad al integrar la simulación basada en la física con modelos basados en datos. Lo que antes requería un gran esfuerzo manual ahora se puede realizar de forma mucho más eficiente, acelerando todo el ciclo de vida de la ingeniería digital y redefiniendo cómo se diseñan y validan los sistemas inteligentes definidos por software”.
Dyad implementará la IA para la ciencia en el mundo real
La IA de propósito general no puede garantizar que un modelo obedezca las leyes de la física. En ingeniería física, un error no es un fallo que se pueda corregir; es el colapso de un puente o el incendio de una batería. Esta ha sido la barrera que ha impedido que la IA desempeñe un papel significativo en la ingeniería de hardware, hasta ahora. En recientes pruebas comparativas de agentes para el modelado de procesos químicos, los sistemas LLM generales como Codex, Claude Code (Opus) y Gemini apenas completaron la configuración inicial. Dyad automatizó casi por completo todo el proceso de creación de controladores predictivos de modelos para optimizar el rendimiento de una planta química, una tarea que normalmente llevaría semanas.
“Se está produciendo una transformación radical en el software de diseño de sistemas de ingeniería, y Dyad está a la vanguardia. Las generaciones anteriores de herramientas no ofrecen la productividad ni la integración prometidas para aprovechar al máximo la IA. Con Dyad, puede modelar la física, desarrollar algoritmos de control con generación automática de código y crear gemelos digitales y modelos sustitutos precisos para el desarrollo rápido de modelos de inferencia de aprendizaje profundo, todo ello gracias a la IA. Dyad opera donde la física se encuentra con el análisis, ¡y los clientes y los accionistas salen ganando!“, afirmó David Joyce, exconsejero delegado de GE Aviation y vicepresidente de GE.
El lenguaje de modelado de Dyad está diseñado específicamente para que los agentes de IA lo comprendan fácilmente. Su lógica fundamental se basa en las leyes de la física, lo que permite a sus agentes razonar sobre cómo se mueven los fluidos a través de las máquinas, cómo la velocidad del viento y la temperatura afectan a los componentes y cómo fuerzas fundamentales como la gravedad dan forma al diseño. Esto produce modelos físicamente válidos en los que los ingenieros pueden confiar. Por ejemplo, en colaboración con Binnies, una empresa con 100 años de experiencia en la gestión del agua, y Williams Grand Prix Technologies,JuliaHub desarrolló un gemelo digital basado en SciML que utiliza solo cuatro entradas de sensores para predecir fallos en las bombas de los sistemas de distribución de agua con una precisión superior al 90 %.
“Dyad representa un cambio radical para el sector del agua, permitiendo pasar de operaciones reactivas a una toma de decisiones predictiva a nivel de sistema”, afirmó Tom Ray, director de Productos y Servicios Digitales (Gemelos Digitales e IA) de Binnies. “Tiene el potencial de transformar la forma en que las empresas modelan la complejidad del mundo real, predicen fallos y optimizan el rendimiento a diario“.
Únase a nosotros para el evento de lanzamiento Dyad 3.0
Dyad 3.0 se presentará oficialmente en un evento en vivo el próximo mes, el 19 de mayo. Únase a nosotros para ver demostraciones del producto en directo y escuchar a nuestros clientes sobre cómo utilizan Dyad en diversos sectores, desde la industria aeroespacial y la climatización hasta los servicios públicos y la robótica.
Para obtener más información y consultas de los medios, contacte con: [email protected]
Acerca de JuliaHub
JuliaHub es líder en IA científica y su misión es brindar a quienes abordan los desafíos científicos y técnicos más complejos del mundo herramientas de vanguardia basadas en IA en un entorno seguro y sin interrupciones. La empresa fue fundada en 2015 por los creadores de Julia, el lenguaje de código abierto de alto rendimiento desarrollado en el MIT y utilizado actualmente por más de un millón de desarrolladores en todo el mundo. JuliaHub combina computación matemática avanzada y experiencia en aprendizaje automático para habilitar técnicas de aprendizaje automático científico (SciML), soluciones de gemelos digitales y modelado y simulación de última generación en los sectores aeroespacial, automotriz y otros sectores industriales.
