Apply Ventures is a venture capital firm that invests in early-stage companies building with and for artificial intelligence. The firm focuses on startups that apply AI to solve concrete problems in industries such as healthcare, robotics, enterprise software, and climate, as well as those developing core AI infrastructure. Apply Ventures typically leads or co-leads seed and Series A rounds, providing capital, operational support, and access to a network of technical advisors and corporate partners.
Founded in 2021 by former operators and engineers from technology companies, Apply Ventures was established to bridge the gap between academic AI research and commercial deployment. The firm's partners have backgrounds in Machine learning systems, Large language model development, and product management, which informs their due diligence and portfolio guidance. As of 2025, the firm manages over $400 million in assets across two funds and has backed more than 30 startups.
Investment Thesis
Apply Ventures' thesis centers on the idea that AI's economic value will accrue not only to model builders but also to companies that integrate Generative AI into workflows, data pipelines, and user interfaces. The firm categorizes its investments into three pillars: applied AI, AI infrastructure, and AI-native tools. In applied AI, it targets sectors with high data density and clear regulatory or operational pain points, such as Intuitive Surgical-adjacent surgical robotics and TomTom-style geospatial analytics. In infrastructure, it supports companies building specialized chips, like those from Groq or SambaNova, and optimization software for Model Pruning and efficient inference.
The firm also invests in foundational research spinouts, often collaborating with academic labs such as Stanford AI Lab and MIT CSAIL. It has funded teams working on Residual Network (ResNet) improvements, Curriculum Learning strategies, and novel Loss Functions for reinforcement learning. Apply Ventures prefers founders with deep technical expertise, often hiring from OpenAI, Anthropic, or Google DeepMind, and it emphasizes measurable product-market fit over speculative research.
Portfolio and Notable Investments
Apply Ventures' portfolio includes a range of companies across the AI stack. In infrastructure, it has invested in a startup developing AWS Trainium-compatible compilers and another creating Graphcore-like IPU alternatives for edge devices. In applied AI, its holdings include a healthcare analytics firm using U-Net architectures for medical imaging, a logistics company applying Sequence-to-Sequence (Seq2Seq) models for route optimization, and a legal tech startup leveraging Top-P (Nucleus) Sampling for document generation.
One notable investment is in a robotics company that uses Cross-Attention mechanisms to fuse vision and language for warehouse automation, competing with players like Figure AI and Sanctuary AI. Another is a fintech startup that applies Batch Normalization and Gradient Clipping to stabilize fraud detection models. The firm has also backed a climate tech company using Data Augmentation to improve satellite-based carbon monitoring, and a developer tools firm that builds Beam Search optimizers for code generation.
Team and Advisors
Apply Ventures is led by managing partners with prior experience at major technology firms. One partner previously led AI product teams at Amazon Web Services, another co-founded a Neural network hardware startup acquired by AMD, and a third served as a research scientist at Xerox PARC. The team includes former engineers from Apple and Samsung Electronics, providing expertise in edge deployment and mobile AI.
The firm maintains a network of technical advisors, including professors from University of Toronto and Carnegie Mellon University, as well as industry veterans like Jakob Uszkoreit, co-inventor of the Transformer (architecture) architecture, and Lukasz Kaiser, known for work on Multi-Head Attention. Advisors participate in quarterly technical reviews and help portfolio companies with model architecture decisions, Learning Rate Scheduling tuning, and Temperature Scaling for production systems.
Approach and Support
Apply Ventures takes a hands-on approach, assigning a dedicated partner to each portfolio company. The firm offers a "technical playbook" that covers Weight Initialization best practices, Dropout strategies, and Positional Encoding choices for custom models. It also provides access to a shared GPU cluster, legal templates for IP protection, and recruiting support for AI engineering roles.
The firm hosts an annual summit where portfolio companies present to corporate partners from Intel, Qualcomm, and Arm Holdings. It also runs a residency program for PhD students from BAIR (Berkeley AI Research) and University of Oxford, offering stipends and mentorship. Apply Ventures publishes an annual report on AI investment trends, which has been cited in industry analyses.
Impact and Future Outlook
Apply Ventures has contributed to the growth of the AI startup ecosystem by funding companies that might otherwise struggle to secure capital due to technical risk. Its portfolio companies have collectively raised over $1.5 billion in follow-on funding. The firm is expanding its focus to include Deep learning for scientific discovery, such as protein folding and materials design, and is exploring investments in Neural network interpretability tools.
As of 2025, Apply Ventures is raising a third fund targeting $250 million, with a planned emphasis on Generative AI applications in regulated industries. The firm remains committed to its thesis that sustainable AI companies require both cutting-edge research and pragmatic engineering, a balance it seeks to foster in every investment.