An open foundation-model company focused on coding and long-horizon software tasks
Build and release foundation models from scratch, prioritizing coding and software tasks as a route toward AGI. Differentiate through reinforcement learning, engineering efficiency, open weights, and genuinely open research rather than only publishing model weights.
online B2B Deep Tech / Infrastructure
From Latent Space — The AI Frontier: from open weights to open research — Eiso Kant, Poolside AI at 02:43
Problem: A small number of companies could otherwise control advanced intelligence, while model-building research is difficult to reproduce and model development cycles are slow, expensive, and engineering-intensive.
For: Companies, countries, and individuals that want a choice of trusted foundation-model providers, especially developers and organizations using models for coding, knowledge work, and long-horizon software tasks.
Examples
- Sourced: Eiso Kant's earlier fully open-source company spent four or five years and $12 million building language models for code before ultimately failing, although the later success of language models served as a vindication of the underlying direction.
Behind this: 14 build steps · 7 tools and how each is used · how to validate demand · 3 more real examples · 7 things the video never answers.
Other takes on AI model development
- Retrain small open-source language models for long-horizon agentic workloads and sell efficient inference and on-device deployments
- AI simulation company that models human behavior and lets organizations test decisions against simulated populations
- Enterprise context and training-data infrastructure for AI applications
- In-database semantic analytics using large database models
- A domain-specific AI product built around private data, specialized models, and purpose-built workflows
- Continually learning AI model provider for enterprises