Ridgeline

Ridgeline is attending the Tough Tech Summit in Boston on October 28—and this is the event if you're serious about understanding where hard tech is headed. This is a power-packed day focused on the technologies that actually move the needle: energy, semiconductors, advanced materials, climate, biotech. This is about infrastructure that's enabling the next wave of breakthroughs—as highlighted by The Engine, Breakthrough Energy Ventures, DCVC, Engine Ventures, Playground Global, RA Capital Management, SOSV, and others pushing this space forward. If you're building in tough tech, investing in it, or just trying to figure out why it matters—we look forward to seeing you there. 📅 Tuesday, October 28 📍 Hotel Commonwealth, Boston, MA 🔗 Event link: https://lnkd.in/g-Cm_6eu #ToughTechWeek2025 #ToughTech #Robotics #AdvancedSystems #AI #Energy #Space #Manufacturing

Why Earth Observation Data Just Became AI's Next Major Modality 🌎 Ridgeline backed LGND AI, Inc., a platform that brings large language models' power to geospatial data. Founders Dan Hammer (ex-NASA, Ode), Nathaniel Manning (co-founder Kettle), and Bruno Sanchez-Andrade Nuño (ex-Mapbox Chief Scientist) are solving a critical unlock: making Earth observation as searchable as Google made the web. Here's why this matters now: – The scale is staggering. While language models train on ~1 petabyte of text, open geospatial imagery exceeds 200 petabytes. That's an untapped goldmine. The data has existed for decades—satellites have been streaming it for 20+ years. What's changed is that transformers can finally extract signal from it. – The unlock is speed and cost. Today, answering "Which dams were retrofitted in the last decade?" costs a quarter-million dollars and months of engineering work. With LGND, an analyst runs a query. What took 6+ months becomes minutes. That changes which problems are worth solving. – It's the next platform shift. Language expanded from keywords to embeddings. Maps are moving from tiles to geo-embeddings—a new way for AI systems to reason over the physical world. This isn't incremental. – Real applications: Wildfire insurance (which properties are fully surrounded by trees?), supply chain (find grain silos with red roofs in Brazil), flood modeling, real estate, and consumer travel—"show me beaches without seasonal seaweed." The use cases are practically infinite. We partnered with LGND because this technology will reshape how government, enterprise, and eventually consumers understand Earth's most pressing challenges. The company is modernizing a foundational sector that is still organized around expensive, manual workflows. Read our full conversation with Dan and Nat: https://lnkd.in/gRTQGFVM #AI #EarthObservation #Geospatial #VentureCapital #Climatetech #Startups #RidgelineBacked

Anu Adedoyin Adasolum just published a piece in Semafor that captures exactly why we backed Sabi: they're building the infrastructure that lets African producers participate in global supply chains on their own terms. DR Congo supplies roughly three-quarters of the world's cobalt. Small-scale mining operations control up to 75% of supply in some regions. But these producers can't access global markets directly because they can't prove provenance digitally—so intermediaries extract most of the value. The EU now mandates digital due diligence and verifiable sourcing data. Other buyers will follow. Sabi built TRACE to solve this with digital infrastructure that proves legitimacy at every step—from mine to port. When a shipment has verified records, financing costs drop, markets open, and small-scale operators can participate without intermediaries taking their margin. Producers gain predictable offtake, governments capture revenues, communities see safer work. This is infrastructure that unlocks billions in trade that's currently trapped by opacity. When you can digitally prove what a commodity is and where it came from, you fundamentally change the economics of how Africa participates in global supply chains. Read the full article: https://lnkd.in/d6x7-67H #Traceability #SupplyChain #Africa #Commodities #RidgelineBacked Ryan Clinton

Most enterprise knowledge is trapped in documents that nobody can find when they actually need them. Implicit just launched KnowledgeOS to fix that—and they're doing something no one else in enterprise AI is doing: offering it as a true freemium product with no time limits. These problems exist across enterprises: an aircraft mechanic needs to troubleshoot a grounded plane but the repair manual doesn't account for this specific configuration. A cybersecurity analyst is responding to an incident and needs to connect CVE data with internal engineering notes. A healthcare worker needs compliant guidance but policies are scattered across multiple systems. In every case, the knowledge exists—it's just functionally inaccessible when it matters most. KnowledgeOS turns unstructured documents—operating manuals, procedures, transcripts, tickets, policies—into a dynamic knowledge graph that doesn't just store content but understands and applies it in context. And unlike typical 14-day trials, users can upload documents into a private, secure environment and explore indefinitely. John K. and team are making enterprise knowledge actually accessible when people need it most. Learn more: https://lnkd.in/gYyhRh2j #EnterpriseAI #KnowledgeManagement #AI #RidgelineBacked Andrew McMahon

Everyone's racing to build AI agents. But here's what most companies are missing: agents that actually know what they're doing before they touch production systems. AMESA (formerly Composabl) just announced their rebrand and expanded platform—and they're tackling the practice problem. While other platforms focus on connecting agents, AMESA gives agents a proving ground where they can learn, fail safely, and gain expertise through simulation and feedback before deployment. Think about what happens when an AI agent makes a mistake in a chemical plant, manages energy distribution across a power grid, or controls warehouse robotics during peak season. Enterprises don't need more agents—they need agents that perform reliably in complex, high-stakes environments where errors cost millions or put safety at risk. AMESA's approach comes from years of proven work in industrial automation, where mistakes are expensive and reliability isn't optional. Their platform lets agents train against digital twins and real-world complexity, then links performance directly to operational metrics like yield, energy efficiency, and downtime. The gap between building an agent and deploying one that performs is where most enterprise AI initiatives stall. AMESA bridges that gap with infrastructure that treats agent competence as seriously as agent connectivity. Learn more: https://lnkd.in/gX_CZahx #AIagents #EnterpriseAI #Automation #MachineLearning #RidgelineBacked Andrew McMahon Kence Anderson

Harbr Data just became a Built on Databricks partner—and this addresses one of the biggest friction points in enterprise AI: how do you actually share data securely across platforms when compliance can't be compromised? The integration creates private data marketplaces inside existing Databricks infrastructure, so enterprises and governments can discover, access, and share live data products without breaking governance policies. It's permission-based sharing that works across different clouds and regions, with built-in compliance for regulated sectors. Here's what makes this powerful: Organizations are already using this to break down data silos while maintaining strict security controls. Instead of choosing between data accessibility and compliance, you get both. Technical teams can query with SQL, business users can use natural language—all within a governed environment. As AI investments scale, the companies that win will be the ones that can move data quickly and securely across their organizations. This partnership gives enterprises the infrastructure to do exactly that, turning data governance from a blocker into an enabler. Congrats to Anthony Cosgrove MBE and the Harbr team on this milestone. Learn more: https://lnkd.in/gQaQxKNf #DataGovernance #EnterpriseAI #DataSecurity #Databricks #RidgelineBacked Andrew McMahon

Q-CTRL just made TIME's Best Inventions of 2025 list! 🎉 Their quantum technology is now addressing real problems that affect thousands of people every single day as an increasing number of commercial flights are disrupted by GPS jamming and denial. Their Ironstone Opal system leverages quantum sensors, stabilized using software, to provide navigation that is immune to the kinds of interference plaguing commercial aviation, shipping, and defense operations. Congratulations to Michael Biercuk and team for this recognition, and for taking frontier technology and making it immediately useful. DARPA recently awarded Q-CTRL $24.4M for quantum sensors on defense platforms. Lockheed Martin and Airbus are actively exploring quantum navigation. The applications span commercial aviation, shipping, and defense operations. Here is a team can cut through the noise and deliver technology that works in the field, passes military-grade testing, and addresses a problem that costs the industry billions annually. Keep up the great work. Read the press release: https://lnkd.in/gvGxwrax #QuantumTech #DeepTech #Aviation #VentureCapital #RidgelineBacked Ryan Clinton

U.S. Customs & Border Protection (CBP) just chose Altana's Product Passports to power trusted trade—and this is a watershed moment for how compliance actually works at the border. Instead of scrambling to prove compliance when goods arrive, importers can now map their entire supply chain upfront—raw materials to finished product—and get CBP approval before shipments even reach the U.S. Once validated, future shipments reference that Product Passport ID and move through customs faster, just like Global Entry for travelers. The shift from reactive to proactive compliance changes the economics and risk profile for everyone. Companies like BASF, L.L.Bean, and Maersk are already using this to cut delays and prove their supply chains are clean. CBP gets verified data before customs filings instead of after, so they can focus enforcement where it actually matters. This is the infrastructure innovation that makes modern trade policy enforceable at scale. As tariffs and regulations increasingly reach back through entire supply chains, companies need systems that can actually track and verify that complexity. Product Passports are that system. Congrats to Evan Smith and the Altana team on this milestone. Learn more: https://lnkd.in/eQAJSSKn #SupplyChain #TradeTech #GlobalTrade #BorderSecurity #RidgelineBacked Ben Walker

SatVu just partnered with Sanborn Map Company to bring thermal satellite intelligence to the US market. Distribution matters as much as technology: the best satellite data is worthless if it doesn't reach the customers who need it. Sanborn has spent 159 years building relationships with government agencies, utilities, and enterprises across America. SatVu's thermal infrared satellites can detect heat signatures at 3.5 meter resolution—revealing energy loss, infrastructure activity, and environmental changes invisible to conventional imagery. But reaching decision-makers at scale requires trusted channels, not just better sensors. The applications span urban heat mapping for climate resilience, industrial monitoring for energy efficiency, and environmental analysis for resource management. Thermal intelligence works day and night, through clouds and low visibility—filling gaps that optical satellites can't address. SatVu is proving that deep tech scales through strategic go-to-market partnerships, not just constellation expansion. Full production begins as they build out their satellite network with a proven route to enterprise and government customers already in place. Congratulations to Anthony Baker and the SatVu team, and to John Copple at Sanborn Map Company on this partnership. Read more: https://lnkd.in/gpSqkJfr #SpaceTech #Satellites #ClimateResilience #Infrastructure #RidgelineBacked

THOR Industries, Inc. just launched the world's first range-extended electric Class A motorhome, with 450 miles of range, built on Harbinger Motors' medium-duty EV chassis. There's something that stands out: Medium-duty electrification works when you start with the use case, not the technology. Harbinger designed this chassis exclusively for THOR Industries' entire family of companies. That partnership model—building purpose-built platforms for specific commercial applications—is how electric medium-duty vehicles actually scale. Not through one-off conversions, but through deep integration with manufacturers who understand their end users. The RV industry faces the same electrification challenge as delivery fleets, utility trucks, and commercial vehicles: range anxiety kills adoption. THOR Industries and manufacturer Entegra Coach addressed it by combining 105 miles of pure electric range with a low-emission range extender. The center-mounted 140-kWh battery pack improves handling and reduces sway—solving a real safety issue in large motorhomes. Full production begins in 2026, with rental fleet testing through THL to gather real-world feedback before scale manufacturing. Congratulations to John Henry Harris and the team at Harbinger Motors, THOR Industries, Entegra Coach, and to Bob Martin and Ken Walters on this milestone. Read the press release (THOR): https://lnkd.in/gn3bynDp Coverage in EV Report: https://lnkd.in/g6D5ZimN #ElectricVehicles #CommercialEV #RVIndustry #Sustainability #RidgelineBacked Ryan Clinton