- Paris-based startup Poolside is in talks to raise over $400 million in funding.
- Bain Capital Ventures and DST are expected to co-lead the funding round, valuing Poolside at $2 billion.
- Led by CEO Jason Warner, Poolside focuses on developing a language model optimized for coding tasks.
- The company uses reinforcement learning from code execution feedback to enhance its model’s capabilities.
- Poolside plans to build a custom AI training toolkit to manage datasets and automate tasks for developing LLMs.
- Synthetic data generation is a key strategy for improving AI training data quality.
- Poolside has secured AI training infrastructure from Australian data center operator Iren, utilizing 504 H100 GPUs.
Main AI News:
Poolside, a Paris-based startup gearing up to launch a significant language model tailored for developers, is reportedly in discussions to secure a substantial funding round. According to sources cited by TechCrunch, the investment could exceed $400 million. Bain Capital Ventures and DST are reportedly in talks to co-lead this round, which would value Poolside at around $2 billion.
Led by CEO Jason Warner, formerly of GitHub, and co-founded by Eiso Kant, known for launching Athenian (acquired by the Linux Foundation), Poolside relocated its headquarters from the U.S. to Paris last August. Earlier, it raised $126 million in initial funding, supported by Bain Capital Ventures and other major backers.
Poolside is focused on developing a language model optimized for coding tasks, leveraging a unique training method called reinforcement learning from code execution feedback. This approach aims to enhance the model’s performance across real-world software projects. Additionally, the company plans to build a specialized AI training toolkit to support researchers in managing datasets and automating tasks related to developing LLMs.
Emphasizing data quality in their development efforts, Poolside utilizes synthetic data generation techniques to bolster its AI capabilities. The startup recently secured a deal with Australian data center operator Iren for AI training infrastructure, including a cluster of 504 H100 GPUs from Nvidia Corp.
Looking ahead, Poolside aims to expand beyond developer productivity tools, envisioning AI models that enable business users to create software and apply advanced reasoning capabilities across various domains.
Conclusion:
Poolside’s ambitious funding efforts and strategic focus on AI language model development highlight its potential to significantly impact the market for developer tools and AI technologies. With a strong emphasis on enhancing productivity and expanding into broader application areas, Poolside is poised to influence how businesses approach software creation and AI-driven decision-making.
