This media is not supported in your browser
VIEW IN TELEGRAM
Spain’s Multiverse Computing has raised €189M (~$217M) in a Series B round led by Bullhound Capital, with backing from HP, Toshiba, and the Spanish government. The company is tackling one of AI’s biggest pain points: cost and scalability of large models.
• Uses tensor networks from quantum physics to compress models by up to 95%.
• Cuts inference costs by 50–80%, with minimal accuracy drop.
• Supports LLaMA, Mistral, and DeepSeek — deployable on phones, PCs, even Raspberry Pi.
• Live on AWS Marketplace, targeting telcos, defense, and AI-at-the-edge use cases.
• Founded in 2019; now Spain’s leading AI scaleup with strong public-private backing.
• €59M of latest round came from the Spanish state, backing sovereign AI infrastructure.
• Compresses open-source models without retraining, saving energy and hardware costs.
• Competes with quantization/pruning — but uses a quantum-inspired framework with better scalability.
• Makes AI more accessible for enterprises outside Big Tech.
Multiverse is betting that compression isn’t a tool—it’s the foundation for the next era of affordable, decentralized AI.
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
Gecko Robotics has entered the unicorn club with a $125M Series D, building wall-climbing robots that inspect critical infrastructure faster, safer, and more accurately than humans.
• Magnetic robots inspect ships, refineries, and power plants — 10× faster than manual checks.
• Capture 1,000× more data per job, feeding into Gecko’s AI platform, Cantilever.
• Reduces risk for workers while boosting maintenance accuracy.
• Used by the U.S. Navy, Abu Dhabi National Oil Co, and NAES.
• Cuts downtime and enables predictive repairs with actionable AI insights.
• Backed by $347M total funding; investors include Founders Fund and Y Combinator.
• Founded in 2016 in a college dorm by Jake Loosararian and Troy Demmer.
• Now valued at $1.25B, expanding across defense, energy, and manufacturing.
Gecko is turning dangerous inspections into a high-speed, high-precision AI process — climbing walls to keep infrastructure upright.
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
Brazilian fintech Meliuz has raised R$180M (~$32.4M) through a follow-on share offering to buy bitcoin, becoming the first public company in the country to formally adopt bitcoin as a treasury asset.
• Funds will be used entirely to purchase bitcoin for long-term holding.
• Marks a bold pivot from cashback and financial services into crypto-native treasury strategy.
• Positions Meliuz as Brazil’s first public “bitcoin treasury” fintech.
• Raised R$180M via primary offering priced at R$7.06/share (~5% discount).
• BTG Pactual acted as sole bookrunner on the offering.
• Shareholder vote in April enabled bitcoin to become a formal reserve asset.
• Move echoes MicroStrategy-style bitcoin accumulation — but in LatAm fintech context.
• Demonstrates growing institutional confidence in bitcoin amid high inflation and currency volatility.
• Could inspire other public firms in Brazil to explore crypto as an alternative asset class.
• Founded in 2011, Meliuz started with cashback and digital finance tools.
• This marks a strategic shift toward crypto-forward treasury and brand positioning.
• Aims to lead in bridging fintech utility with decentralized asset adoption.
Meliuz is betting that bitcoin isn’t just a store of value — it’s a new financial identity for the next generation of LatAm fintech.
Please open Telegram to view this post
VIEW IN TELEGRAM
Meta has acquired a 49% non-voting stake in Scale AI for $14.3 billion, valuing the data-labeling powerhouse at ~$29B — and bringing on its co-founder and CEO Alexandr Wang to lead Meta’s “superintelligence” initiative.
• Wang will head Meta’s new AGI division, reporting directly to Mark Zuckerberg.
• The group will focus on developing artificial general intelligence with access to Scale’s massive data infrastructure.
• Wang remains on Scale’s board, while CSO Jason Droege becomes interim CEO.
• Meta’s stake is non-voting — ensuring Scale’s operational independence.
• The structure avoids regulatory pushback while securing data access and leadership.
• Scale remains a core vendor to OpenAI, Microsoft, Google, and U.S. defense — for now.
• Comes after disappointing traction with Meta’s Llama models in the enterprise market.
• Zuckerberg has been on an AI talent spree — offering up to 9-figure comp to top researchers.
• The deal positions Meta to vertically integrate data labeling and model training.
• Scale AI was founded in 2016 by Alexandr Wang and Lucy Guo.
• Went from defense contracts to powering AI pipelines for Silicon Valley giants.
• Now at the center of Meta’s most ambitious AI play yet.
Meta didn’t just invest in Scale — it invested in its architect. Wang’s move signals a new phase of AGI competition, where infrastructure and intelligence are built in-house.
Please open Telegram to view this post
VIEW IN TELEGRAM
Media is too big
VIEW IN TELEGRAM
Munich-based Proxima Fusion has closed a €130M Series A to develop the world’s first commercial stellarator-based fusion power plant, offering a radically stable path to net-energy fusion in the 2030s.
• Uses stellarators, which confine plasma with external magnets — unlike Tokamaks, no internal currents = better stability.
• Combines quasi-isodynamic designs with AI and high-temperature superconductors (HTS) for compact, scalable machines.
• Based on research from the Max Planck Institute’s Wendelstein 7-X, the world’s most advanced stellarator.
• 2027: Targeting “Model Coil” milestone — a full-scale magnet testbed.
• 2031: Net-energy prototype named Alpha to be operational.
• 2030s: Launch of Stellaris, the first grid-ready fusion power plant.
• €130M Series A led by Redalpine, UVC Partners, Atlantic Labs, and Earlybird.
• Europe’s largest private fusion round to date; brings total raised to €185M+.
• German government support signals rising momentum behind deeptech energy bets.
• Technical hurdles: precision magnet manufacturing, tritium breeding, and power conversion.
• Fusion remains a long game — capital-intensive with multi-decade ROI timelines.
• But stellarators may offer a more commercially viable and steady-state alternative.
Proxima isn’t chasing the next fusion breakthrough — it’s engineering the one we already have into reality.
Please open Telegram to view this post
VIEW IN TELEGRAM