Two AI teams working on LoRA continuous learning: One is http://river.ai, founded by CEO Igor Babuschkin (former co-founder of xAI, previously worked on generative modeling and reinforcement learning at DeepMind, and large-scale training at OpenAI). The core team comes from xAI, Tesla, with backgrounds in DeepMind/OpenAI. Funding: Seed + Series A totaling $1.1 billion. Using your data and your own GPU, you can train a LoRA model that continuously learns from your experiences. The other is a LoRA company invested in by Meituan DragonBall called Mindverse (funding exceeds $50 million). They focus on continuous learning through LoRA. Their approach is based on fine-tuning open-source models (they’ve developed their own LoRA RL framework). Their early product was an independent app (Macaron app), a personal agent for the lifestyle space. However, their main disclosed revenue doesn’t seem to come from the app but rather from their RL-for-LoRA model fine-tuning cloud service. The app experience is pretty average, but its early positioning was similar to agent products like Muse. It seems the company’s focus has shifted to RL cloud services. The team appears to lean more toward the technical side rather than being a consumer-facing team. Company/Product: https://macaron.im (C-end Personal Agent 'Macaron') Research Lab Mind Lab: https://macaron.im/mindlab