Participated · Series B · Aug 2026

Flagler Health builds an AI-native platform that integrates directly into provider EMRs and orchestrates the full musculoskeletal patient lifecycle across triage, care management, billing, and patient communications. The platform is trained on one of the largest MSK patient data sets and is designed to run seamlessly in the background with no new logins or workflow changes for clinicians. In under three years, Flagler has scaled to thousands of providers across more than 36 states and reports metrics including an average $164,000 in additional annual revenue per provider and 87% of patients reporting improvement in pain, mood, sleep, or mobility. The product has been deployed across diverse MSK practice types and EMR environments and was developed by a founding team of physicians and healthcare operators. Headquartered in New York City, Flagler positions itself as the AI-native backbone for musculoskeletal clinics and aims to expand to every MSK practice in America.

Participated · Series B · Feb 2026

Ownwell modernizes the traditionally opaque property-tax appeal process by combining local tax expertise with proprietary technology that automates complex steps and analyzes millions of property records. The company manages appeals end-to-end with no upfront fees and charges a 25% contingency only when savings are achieved. To date, it has processed more than one million appeals, delivering over $400 million in total savings for property owners and averaging $774 in annual savings per customer with an 86% success rate. Ownwell’s core offering now spans Texas, Illinois, Florida, Georgia, California, Washington, and New York, and the firm has launched a National Appeals Packet that uses AI-generated data to let homeowners nationwide file their own appeals. Beyond tax appeals, it also offers tax-bill management, mortgage, insurance, energy-saving, and other bill-reduction services to lower overall homeownership costs. The company intends to use new capital to accelerate product development and support nationwide expansion so it can help “millions more” homeowners combat rising costs. CEO and founder Colton Pace positions Ownwell as a “done-for-you” solution that levels the playing field between individual homeowners and large real-estate investors.

Participated · Series B · May 2025

TurbineOne builds an edge-first AI platform, the Frontline Perception System (FPS), which runs directly on battlefield sensors and everyday military gear without requiring cloud access or custom hardware. The platform is designed by veterans for austere conditions and is deployed across the U.S. Department of Defense for missions including drone warfare, base security, and targeting operations. The company was founded in 2021 and is headquartered in San Francisco. TurbineOne originated through work with innovation organizations such as the Defense Innovation Unit and AFWERX and is transitioning to deliver operational capabilities for the Army, Navy, and Air Force. Its software emphasizes speed, simplicity, and reliability to provide decision advantage for warfighters. The new funding will support global deployments, hardened self-service products, and hiring across engineering and operations. TurbineOne builds the Frontline Perception System (FPS), an edge-first, no-code, no-cloud ML platform that automates threat recognition and refines raw data feeds for decision-makers in constrained battlefield conditions. The software is designed to operate on a wide range of systems, works offline in low-connectivity and limited-bandwidth environments, and integrates data from many human-understandable sources. TurbineOne says FPS has operated on U.S. military autonomous systems and in tactical operations centers, with demonstrated use cases across Navy, Army, and Special Operations units. The company positions its product as user-friendly so warfighters can retrain and operate ML models themselves without cloud reliance. Financially, TurbineOne announced a $15 million Series A to support expansion in national security markets as the Defense Department adopts ML solutions. TurbineOne builds the Frontline Perception System, a wearable and edge-deployable platform that uses machine learning on edge hardware to deliver situational awareness without relying on internet connectivity. The system breaks down data silos and provides real-time teammate location, smart alerts for unseen dangers, and a marketplace-like model for third-party perception models. TurbineOne is expanding the product beyond body-worn systems to include robots and drones, and will power autonomous robots to geolocate Foreign Object Debris (FOD) on military airfields under a Department of Defense prototyping contract at Joint Base Pearl Harbor-Hickam with Siemens Government Technologies. The company works with the U.S. Navy and Air Force and counts Siemens among its commercial partners. TurbineOne announced a $3 million seed raise to expand its engineering team and accelerate product delivery. The company is based in San Francisco and was incubated by Sweat Equity Ventures; its leadership includes co-founders Ian Kalin and Matt Amacker.

Participated · Series B · Jul 2024

HerculesAI, formerly Zero Systems, has been working since 2017 on automating professional services and building large language models. The company develops models that perform three core functions: intelligent data extraction, data transformation, and data verification, and it layers a multi-agent system on top to automate decision-making workflows. HerculesAI began building small models (about 0.5–2 billion parameters) around 2020 and has run them on edge devices for compliance purposes. Its regulated-industry customer base includes legal, insurance, and financial services; the company counts 30% of the top 100 U.S. law firms and Fortune 500 clients such as Mercer, Standard & Poor's, and State Farm. The business reports 4x growth over the last year and currently has around 75 employees. Management plans to remain lean and invest the new capital in refining internal processes and further automation, using its own products to scale. Zero Systems builds an AI-driven automation platform that identifies work product across lawyers' devices, classifies it by client and project, and routes documents, emails and messages into adjacent systems like document management and time tracking. The solution is installed on customers' premises and processes data at the edge for security and to meet customer requirements. The company focuses on high-value, cognitive processes—governance filing, automated time capture and security controls—rather than bulk data movement. Zero launched seven years ago and released its product about two and a half years ago, and has expanded its target beyond law firms into financial services and consulting. It has traction with the AmLaw 100: 11 customers are live and another 10 are piloting. Headcount grew from roughly 25 employees at the start of last year to almost 80 by year-end as it scales product and go-to-market.

Participated · Series B · Mar 2024

Pandion is a Seattle-based parcel network that manages package journeys for e-commerce brands, picking up at fulfillment centers, sorting through its connected sortation centers, and delivering via a network of 500,000+ final-mile drivers. Founded in 2020 by Amazon Air founder Scott Ruffin, the company leverages machine learning and a proprietary universal label to optimize delivery decisions and lower prices versus national and regional carriers. Pandion currently reaches more than 80% of U.S. homes and operates sortation centers near Philadelphia, Dallas, Los Angeles, Chicago, and Atlanta, with capacity to deliver millions of packages. The company provides brands instant access to delivery options and faster transit times for customers such as Saks Fifth Avenue. Pandion plans to use new funding to build additional technology offerings, expand geographic reach, and increase delivery speed. Recent leadership hires and a board addition are intended to support this next phase of growth. Pandion operates a parcel network purpose-built for e-commerce, using machine learning to more efficiently transport packages and save retailers and consumers time and money. The company is opening its first 150,000 square foot sortation facility in Quakertown, Pennsylvania this fall, a site leased through Cabot Properties that can sort and deliver millions of packages per month. Pandion plans to hire more than 80 people at the Quakertown facility and is scaling its technology and sortation-center teams ahead of the holiday shopping season. The team includes executives with experience running Amazon fulfillment, Walmart sortation, FedEx and UPS operations, and machine-learning work for companies like Cruise, Stitch Fix, and Tesla. The company has assembled an advisory board of former senior operations and C-suite executives from major logistics and retail firms. Pandion emerged from stealth in February 2021 and was founded by Scott Ruffin, an Amazon Air and Walmart e-commerce transportation veteran. Pandion is a parcel network purpose-built for e-commerce, offering next-generation delivery management software that uses machine learning to avoid shipping delays and optimize delivery networks. The platform aims to enable consistent one- and two-day shipping and to scale capacity up and down to meet peak-period demand, addressing constraints in legacy parcel providers. Pandion positions itself to help retailers deliver faster, more cost-effective service and to make on-time delivery possible for more retailers regardless of size. Founded in 2020 by Scott Ruffin, the team includes delivery and supply-chain executives from Amazon, Walmart, FedEx, Macy’s, HP, Accenture, and Deloitte. The company emerged from stealth and plans to use its seed proceeds to grow the team and expand ongoing customer pilots. Pandion says it is a long-term scalable solution to e-commerce capacity problems exposed during peak seasons.