VCs Flood AI Infrastructure with Cash

The venture capital flood into AI isn't just about building bigger models; it's about creating the foundational layers that make AI practical in the real world. Encord's latest $60M round, led by Wellington Management, brings its total funding to $110 million and highlights a crucial industry bottleneck: data readiness for "physical AI." This specialized field moves AI from digital spaces to tangible applications like robotics and autonomous vehicles by processing complex, real-world sensor data like LiDAR and video feeds. Encord's platform is designed to manage the lifecycle of this complex multimodal data, a challenge that legacy data systems weren't built for. The company serves over 300 teams, including major players like Woven by Toyota and Skydio, and has seen a tenfold increase in revenue from its physical AI customers over the last year. This growth underscores the market's shift from focusing on model size to ensuring the quality and curation of the data that feeds these sophisticated systems. The trend of "AI-native" solutions extends to cybersecurity, where startups like Gambit Security are attracting significant capital. Gambit emerged from stealth with $61 million from investors including Spark Capital and Kleiner Perkins to build a resilience platform that automates incident recovery. Traditional security tools struggle with the dynamic nature of AI, so AI-native platforms are being developed to address issues like prompt injections and data leakage in real-time. Meanwhile, Akave's $6.65M raise points to another critical infrastructure piece: decentralized data storage. By using a dedicated Avalanche L1 blockchain, Akave offers S3-compatible storage with on-chain verifiability and no vendor lock-in, aiming to give enterprises more control over their data for AI workloads. This approach challenges the dominance of major cloud providers by offering predictable pricing and eliminating egress fees. For new ML engineers, this infrastructure boom signals a clear career trajectory. Proficiency in MLOps is now essential, requiring skills in CI/CD pipelines, containerization with Docker and Kubernetes, and workflow orchestration tools like Airflow. Expertise in cloud platforms such as AWS SageMaker and Google Vertex AI, along with data versioning tools like DVC, is also in high demand. Beyond tooling, top companies are seeking engineers with strong software engineering fundamentals, including a deep understanding of data structures and algorithms. Demonstrating experience with the full machine learning lifecycle—from data preparation and feature engineering to model deployment and monitoring—is crucial for landing competitive roles. Portfolio projects that showcase these production-oriented skills will stand out more than those confined to notebooks.